# UplinkSwitch

> One API key and one base URL for Claude, GPT, Gemini, Grok, DeepSeek, Qwen, Kimi, GLM and MiniMax chat models, priced below the developers' list prices, plus image, video, speech and 3D models (Nano Banana, GPT Image, Seedream, FLUX, Veo, Seedance, Kling, Wan, ElevenLabs, Tripo). Prepaid credit, no subscription.

- Base URL: https://api.uplinkswitch.com/v1 (OpenAI, Anthropic and Gemini SDK formats)
- Auth: `Authorization: Bearer sk-uplink-live-...` (or `x-api-key` / `x-goog-api-key`, as each SDK sends it)
- Keys: https://uplinkswitch.ai/console
- OpenAPI 3.1 document: https://api.uplinkswitch.com/openapi.json (every customer route, request and response shape, and the x-uplink-* headers)

## Every model and price, one line each

- [llms-full.txt](https://uplinkswitch.ai/llms-full.txt): all chat and media models with our price and the developer's, from the live catalogs

## Markdown

- Every page under /models, /pricing, /compare, /blog, /guides, /uncensored, /red-teaming, /pricing-index and /proof has a markdown twin at the same path plus .md (https://uplinkswitch.ai/models/claude-opus-5-5.md, https://uplinkswitch.ai/guides/claude-code.md), or send `Accept: text/markdown` to the page URL. Same live prices as the HTML, no scripts.

## Prices (JSON, no key needed)

- [Chat models](https://api.uplinkswitch.com/v1/catalog): customer price and developer list price per million tokens; the same list carries the embedding, rerank, moderation and transcription models (`type`), the server tools and their prices (`server_tools`), file parsing (`file_parsing`) and the auto router tiers (`auto_router`)
- [Media models](https://api.uplinkswitch.com/v1/catalog/media): settings and price per image, second or generation
- [Price index](https://uplinkswitch.ai/pricing-index): every listed chat model with a developer list price, our price next to it, as [JSON](https://uplinkswitch.ai/pricing-index.json) and [CSV](https://uplinkswitch.ai/pricing-index.csv)
- [AI catalog](https://uplinkswitch.ai/ai-catalog.json): every chat and media model in one document, with our price, the developer's list price, the endpoints and formats it answers, and its page
- [API discovery](https://uplinkswitch.ai/.well-known/api-catalog): RFC 9727 Linkset pointing at the OpenAPI document, the catalogs, the price index, proof and status
- [MCP server](https://api.uplinkswitch.com/mcp): Streamable HTTP, stateless. Tools list_models, list_media_models, estimate and proof need no key; chat is billed to the `Authorization: Bearer sk-uplink-live-...` header. Server card at https://api.uplinkswitch.com/mcp/server-card, AI catalog at https://uplinkswitch.ai/.well-known/ai-catalog.json. Setup for Claude Code, Cursor and Cline in their guides.

## Open source

- [github.com/uplinkswitch](https://github.com/uplinkswitch): the Vercel AI SDK provider (`npm install @uplinkswitch/ai-sdk-provider`) and runnable examples for every SDK and tool

## Uncensored and abliterated models (148)

- [Uncensored models on one key](https://uplinkswitch.ai/uncensored): every build below on one key, with the daily real-model check at https://uplinkswitch.ai/proof
- [GLM-5.3 Uncensored](https://uplinkswitch.ai/models/glm-5.3-uncensored): Z.ai's uncensored build (served without a content filter), served as model id glm-5.3-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $0.30 in / $0.90 out per million tokens, prepaid credit, no subscription.
- [GLM-5.3 Abliterated](https://uplinkswitch.ai/models/glm-5.3-abliterated): Z.ai's abliterated build (the refusal direction removed from the weights), served as model id glm-5.3-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $2.25 in / $4.50 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4 Pro Uncensored](https://uplinkswitch.ai/models/deepseek-v4-pro-uncensored): DeepSeek's uncensored build (served without a content filter), served as model id deepseek-v4-pro-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.00 in / $2.00 out per million tokens, prepaid credit, no subscription.
- [GLM-5.3 Flash Abliterated](https://uplinkswitch.ai/models/glm-5.3-flash-abliterated): Z.ai's abliterated build (the refusal direction removed from the weights), served as model id glm-5.3-flash-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $1.76 in / $1.76 out per million tokens, prepaid credit, no subscription.
- [GLM-5.3 Flash Uncensored](https://uplinkswitch.ai/models/glm-5.3-flash-uncensored): Z.ai's uncensored build (served without a content filter), served as model id glm-5.3-flash-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $0.112 in / $0.392 out per million tokens, prepaid credit, no subscription.
- [Kimi K3 Uncensored](https://uplinkswitch.ai/models/kimi-k3-uncensored): Moonshot AI's uncensored build (served without a content filter), served as model id kimi-k3-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $5.745 in / $28.68 out per million tokens, prepaid credit, no subscription.
- [Nemotron 3 Ultra 550B Uncensored](https://uplinkswitch.ai/models/nemotron-3-ultra-550b-a55b-uncensored): NVIDIA's uncensored build (served without a content filter), served as model id nemotron-3-ultra-550b-a55b-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.125 in / $6.60 out per million tokens, prepaid credit, no subscription.
- [Abliterated Model](https://uplinkswitch.ai/models/abliterated-model): Abliteration.ai's abliterated build (the refusal direction removed from the weights), served as model id abliterated-model on the OpenAI, Anthropic and Gemini SDK formats at $2.50 in / $7.50 out per million tokens, prepaid credit, no subscription.
- [Abliterated Model Large](https://uplinkswitch.ai/models/abliterated-model-large): Abliteration.ai's abliterated build (the refusal direction removed from the weights), served as model id abliterated-model-large on the OpenAI, Anthropic and Gemini SDK formats at $7.50 in / $12.50 out per million tokens, prepaid credit, no subscription.
- [Abliterated Large Flash](https://uplinkswitch.ai/models/abliterated-model-large-flash): Abliteration.ai's abliterated build (the refusal direction removed from the weights), served as model id abliterated-model-large-flash on the OpenAI, Anthropic and Gemini SDK formats at $2.50 in / $5.00 out per million tokens, prepaid credit, no subscription.
- [Abliterated Model Large V2](https://uplinkswitch.ai/models/abliterated-model-large-v2): Abliteration.ai's abliterated build (the refusal direction removed from the weights), served as model id abliterated-model-large-v2 on the OpenAI, Anthropic and Gemini SDK formats at $7.50 in / $12.50 out per million tokens, prepaid credit, no subscription.
- [Aion RP Llama 3.1 8B](https://uplinkswitch.ai/models/aion-rp-llama-3.1-8b): Aion Labs's roleplay build (answers in character instead of refusing), served as model id aion-rp-llama-3.1-8b on the OpenAI, Anthropic and Gemini SDK formats at $3.20 in / $6.40 out per million tokens, prepaid credit, no subscription.
- [Amoral Gemma3 27B v2](https://uplinkswitch.ai/models/amoral-gemma3-27b-v2): soob3123's uncensored build (served without a content filter), served as model id amoral-gemma3-27b-v2 on the OpenAI, Anthropic and Gemini SDK formats at $1.05 in / $1.05 out per million tokens, prepaid credit, no subscription.
- [Anubis 70B v1.1](https://uplinkswitch.ai/models/anubis-70b-v1.1): TheDrummer's roleplay build (answers in character instead of refusing), served as model id anubis-70b-v1.1 on the OpenAI, Anthropic and Gemini SDK formats at $1.085 in / $1.085 out per million tokens, prepaid credit, no subscription.
- [Artemis v1.1](https://uplinkswitch.ai/models/artemis-v1.1): TheDrummer's roleplay build (answers in character instead of refusing), served as model id artemis-v1.1 on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Cu-Mai R1 70B](https://uplinkswitch.ai/models/cu-mai-r1-70b): Steelskull's roleplay build (answers in character instead of refusing), served as model id cu-mai-r1-70b on the OpenAI, Anthropic and Gemini SDK formats at $1.7255 in / $1.7255 out per million tokens, prepaid credit, no subscription.
- [CyberHeabsy](https://uplinkswitch.ai/models/cyberheabsy): heabsy's uncensored build (served without a content filter), served as model id cyberheabsy on the OpenAI, Anthropic and Gemini SDK formats at $5.25 in / $15.75 out per million tokens, prepaid credit, no subscription.
- [Cydonia 24B v4.1](https://uplinkswitch.ai/models/cydonia-24b-v4.1): TheDrummer's roleplay build (answers in character instead of refusing), served as model id cydonia-24b-v4.1 on the OpenAI, Anthropic and Gemini SDK formats at $1.20 in / $2.00 out per million tokens, prepaid credit, no subscription.
- [Cydonia 24B v4.3](https://uplinkswitch.ai/models/cydonia-24b-v4.3): TheDrummer's roleplay build (answers in character instead of refusing), served as model id cydonia-24b-v4.3 on the OpenAI, Anthropic and Gemini SDK formats at $0.484 in / $0.60 out per million tokens, prepaid credit, no subscription.
- [DansPE 1.3 24B](https://uplinkswitch.ai/models/dans-pe-1.3-24b): PocketDoc's roleplay build (answers in character instead of refusing), served as model id dans-pe-1.3-24b on the OpenAI, Anthropic and Gemini SDK formats at $2.908 in / $11.632 out per million tokens, prepaid credit, no subscription.
- [DeepSeek R1 Distill Llama 70B Abliterated](https://uplinkswitch.ai/models/deepseek-r1-distill-llama-70b-abliterated): huihui-ai's abliterated build (the refusal direction removed from the weights), served as model id deepseek-r1-distill-llama-70b-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $2.80 in / $2.80 out per million tokens, prepaid credit, no subscription.
- [DeepSeek R1 Distill Qwen 32B Abliterated](https://uplinkswitch.ai/models/deepseek-r1-distill-qwen-32b-abliterated): huihui-ai's abliterated build (the refusal direction removed from the weights), served as model id deepseek-r1-distill-qwen-32b-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $5.60 in / $5.60 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash 0910 Uncensored](https://uplinkswitch.ai/models/deepseek-v4.1-flash-0910-uncensored): DeepSeek's uncensored build (served without a content filter), served as model id deepseek-v4.1-flash-0910-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $0.825 in / $3.30 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash 0910 Uncensored Thinking](https://uplinkswitch.ai/models/deepseek-v4.1-flash-0910-uncensored-thinking): DeepSeek's uncensored build (served without a content filter), served as model id deepseek-v4.1-flash-0910-uncensored-thinking on the OpenAI, Anthropic and Gemini SDK formats at $0.825 in / $3.30 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash 1017 Uncensored](https://uplinkswitch.ai/models/deepseek-v4.1-flash-1017-uncensored): DeepSeek's uncensored build (served without a content filter), served as model id deepseek-v4.1-flash-1017-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $0.825 in / $3.30 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash 1017 Uncensored Thinking](https://uplinkswitch.ai/models/deepseek-v4.1-flash-1017-uncensored-thinking): DeepSeek's uncensored build (served without a content filter), served as model id deepseek-v4.1-flash-1017-uncensored-thinking on the OpenAI, Anthropic and Gemini SDK formats at $0.825 in / $3.30 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash Abliterated](https://uplinkswitch.ai/models/deepseek-v4.1-flash-abliterated): DeepSeek's abliterated build (the refusal direction removed from the weights), served as model id deepseek-v4.1-flash-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $0.875 in / $1.75 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash Abliterated Fast](https://uplinkswitch.ai/models/deepseek-v4.1-flash-abliterated-fast): DeepSeek's abliterated build (the refusal direction removed from the weights), served as model id deepseek-v4.1-flash-abliterated-fast on the OpenAI, Anthropic and Gemini SDK formats at $1.75 in / $3.50 out per million tokens, prepaid credit, no subscription.
- [Uplink Red Team (Flash)](https://uplinkswitch.ai/models/uplink-red-team-flash): UplinkSwitch's own uncensored build (served without a content filter), served as model id uplink-red-team-flash on the OpenAI, Anthropic and Gemini SDK formats at $0.90 in / $3.00 out per million tokens, prepaid credit, no subscription.
- [DeepSeek V4.1 Flash Uncensored Thinking](https://uplinkswitch.ai/models/deepseek-v4.1-flash-uncensored-thinking): DeepSeek's uncensored build (served without a content filter), served as model id deepseek-v4.1-flash-uncensored-thinking on the OpenAI, Anthropic and Gemini SDK formats at $0.825 in / $3.30 out per million tokens, prepaid credit, no subscription.
- [Dolphin Mistral 24B Venice Edition](https://uplinkswitch.ai/models/dolphin-mistral-24b-venice): Cognitive Computations's uncensored build (served without a content filter), served as model id dolphin-mistral-24b-venice on the OpenAI, Anthropic and Gemini SDK formats at $0.50 in / $2.25 out per million tokens, prepaid credit, no subscription.
- [Euryale 70B v2.2](https://uplinkswitch.ai/models/euryale-70b-v2.2): Sao10K's roleplay build (answers in character instead of refusing), served as model id euryale-70b-v2.2 on the OpenAI, Anthropic and Gemini SDK formats at $1.224 in / $1.428 out per million tokens, prepaid credit, no subscription.
- [Euryale 70B v2.3](https://uplinkswitch.ai/models/euryale-70b-v2.3): Sao10K's roleplay build (answers in character instead of refusing), served as model id euryale-70b-v2.3 on the OpenAI, Anthropic and Gemini SDK formats at $1.972 in / $1.972 out per million tokens, prepaid credit, no subscription.
- [Gemma 31B Heretic Uncensored](https://uplinkswitch.ai/models/gemma-31b-heretic-uncensored): Google's abliterated build (the refusal direction removed from the weights), served as model id gemma-31b-heretic-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.50 in / $8.25 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 12B Starry Ocean](https://uplinkswitch.ai/models/gemma-4-12b-starry-ocean): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-12b-starry-ocean on the OpenAI, Anthropic and Gemini SDK formats at $0.175 in / $0.875 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 12B StationKeeper](https://uplinkswitch.ai/models/gemma-4-12b-station-keeper): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-12b-station-keeper on the OpenAI, Anthropic and Gemini SDK formats at $0.175 in / $0.875 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B ChimeraX](https://uplinkswitch.ai/models/gemma-4-26b-a4b-chimerax): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-chimerax on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B ChimeraX Heretic](https://uplinkswitch.ai/models/gemma-4-26b-a4b-chimerax-heretic): Google (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id gemma-4-26b-a4b-chimerax-heretic on the OpenAI, Anthropic and Gemini SDK formats at $0.484 in / $1.52 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B DarkSoul](https://uplinkswitch.ai/models/gemma-4-26b-a4b-darksoul): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-darksoul on the OpenAI, Anthropic and Gemini SDK formats at $0.484 in / $1.52 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B ChimeraX](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-chimerax): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-chimerax on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B Cybersecurity](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-cybersecurity): Google (community tune)'s uncensored build (served without a content filter), served as model id gemma-4-26b-a4b-it-cybersecurity on the OpenAI, Anthropic and Gemini SDK formats at $0.4224 in / $1.3376 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B DarkSoul](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-darksoul): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-darksoul on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B Luminous](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-luminous): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-luminous on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B MeroMero](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-meromero): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-meromero on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B Moonlight](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-moonlight): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-moonlight on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B Musica](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-musica): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-musica on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B OpusDistill](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-opusdistill): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-opusdistill on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B A4B ShadowSiren](https://uplinkswitch.ai/models/gemma-4-26b-a4b-it-shadowsiren): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-it-shadowsiren on the OpenAI, Anthropic and Gemini SDK formats at $0.48 in / $1.48 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B Luminous Mirror](https://uplinkswitch.ai/models/gemma-4-26b-a4b-luminous-mirror): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-luminous-mirror on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B MeroMero](https://uplinkswitch.ai/models/gemma-4-26b-a4b-meromero): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-meromero on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B Moonlight Dusk](https://uplinkswitch.ai/models/gemma-4-26b-a4b-moonlight-dusk): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-moonlight-dusk on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B Musica](https://uplinkswitch.ai/models/gemma-4-26b-a4b-musica): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-musica on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B Opus Distill](https://uplinkswitch.ai/models/gemma-4-26b-a4b-opus-distill): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-opus-distill on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B Shadow Siren](https://uplinkswitch.ai/models/gemma-4-26b-a4b-shadow-siren): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-26b-a4b-shadow-siren on the OpenAI, Anthropic and Gemini SDK formats at $0.42 in / $1.33 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 26B-A4B Uncensored](https://uplinkswitch.ai/models/gemma-4-26b-a4b-uncensored): Google (community tune)'s uncensored build (served without a content filter), served as model id gemma-4-26b-a4b-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $0.60 in / $2.80 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Agares v1](https://uplinkswitch.ai/models/gemma-4-31b-agares-v1): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-agares-v1 on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Animus V14.1](https://uplinkswitch.ai/models/gemma-4-31b-animus-v14.1): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-animus-v14.1 on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B AssGuard](https://uplinkswitch.ai/models/gemma-4-31b-assguard): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-assguard on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Cognitive Unshackled](https://uplinkswitch.ai/models/gemma-4-31b-cognitive-unshackled): Google (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id gemma-4-31b-cognitive-unshackled on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Dark Gemistry](https://uplinkswitch.ai/models/gemma-4-31b-dark-gemistry): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-dark-gemistry on the OpenAI, Anthropic and Gemini SDK formats at $1.224 in / $1.224 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B DarkIdol](https://uplinkswitch.ai/models/gemma-4-31b-darkidol): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-darkidol on the OpenAI, Anthropic and Gemini SDK formats at $0.40 in / $1.80 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Fabled](https://uplinkswitch.ai/models/gemma-4-31b-fabled): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-fabled on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Garnet V2](https://uplinkswitch.ai/models/gemma-4-31b-garnet-v2): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-garnet-v2 on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Gembrain](https://uplinkswitch.ai/models/gemma-4-31b-gembrain): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-gembrain on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Gembrain Heretic](https://uplinkswitch.ai/models/gemma-4-31b-gembrain-heretic): Google (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id gemma-4-31b-gembrain-heretic on the OpenAI, Anthropic and Gemini SDK formats at $1.224 in / $1.224 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Gembrain X Core](https://uplinkswitch.ai/models/gemma-4-31b-gembrain-x-core): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-gembrain-x-core on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Gemsicle](https://uplinkswitch.ai/models/gemma-4-31b-gemsicle): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-gemsicle on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Isometry](https://uplinkswitch.ai/models/gemma-4-31b-isometry): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-isometry on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Isometry RP](https://uplinkswitch.ai/models/gemma-4-31b-isometry-rp): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-isometry-rp on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B GemBrain Heretic](https://uplinkswitch.ai/models/gemma-4-31b-it-gembrain-heretic): Google (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id gemma-4-31b-it-gembrain-heretic on the OpenAI, Anthropic and Gemini SDK formats at $0.40 in / $1.76 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B K1 v5](https://uplinkswitch.ai/models/gemma-4-31b-k1-v5): Google (community tune)'s uncensored build (served without a content filter), served as model id gemma-4-31b-k1-v5 on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B MeroMero](https://uplinkswitch.ai/models/gemma-4-31b-meromero): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-meromero on the OpenAI, Anthropic and Gemini SDK formats at $0.40 in / $1.76 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B MeroMero v2](https://uplinkswitch.ai/models/gemma-4-31b-meromero-v2): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-meromero-v2 on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Novelist](https://uplinkswitch.ai/models/gemma-4-31b-novelist): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-novelist on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Queen](https://uplinkswitch.ai/models/gemma-4-31b-queen): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-queen on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Scarlet](https://uplinkswitch.ai/models/gemma-4-31b-scarlet): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-scarlet on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.575 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B SDFT Heretic RP](https://uplinkswitch.ai/models/gemma-4-31b-sdft-heretic-rp): Google (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id gemma-4-31b-sdft-heretic-rp on the OpenAI, Anthropic and Gemini SDK formats at $1.224 in / $1.224 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B Split-Untied](https://uplinkswitch.ai/models/gemma-4-31b-split-untied): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-split-untied on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $1.05 out per million tokens, prepaid credit, no subscription.
- [Gemma 4 31B StyleTune](https://uplinkswitch.ai/models/gemma-4-31b-styletune): Google (community tune)'s roleplay build (answers in character instead of refusing), served as model id gemma-4-31b-styletune on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [GLM-4.6 Derestricted v5](https://uplinkswitch.ai/models/glm-4.6-derestricted-v5): Z.ai's abliterated build (the refusal direction removed from the weights), served as model id glm-4.6-derestricted-v5 on the OpenAI, Anthropic and Gemini SDK formats at $1.40 in / $5.25 out per million tokens, prepaid credit, no subscription.
- [GLM-5.3 Flash Apex Abliterated](https://uplinkswitch.ai/models/glm-5.3-flash-apex-abliterated): Z.ai's abliterated build (the refusal direction removed from the weights), served as model id glm-5.3-flash-apex-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $0.80 in / $3.20 out per million tokens, prepaid credit, no subscription.
- [GLM-5.3 Flash Code Uncensored](https://uplinkswitch.ai/models/glm-5.3-flash-code-uncensored): Z.ai's uncensored build (served without a content filter), served as model id glm-5.3-flash-code-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $6.00 in / $6.00 out per million tokens, prepaid credit, no subscription.
- [GLM-5.3 Flash Cybersecurity](https://uplinkswitch.ai/models/glm-5.3-flash-cybersecurity): Z.ai's uncensored build (served without a content filter), served as model id glm-5.3-flash-cybersecurity on the OpenAI, Anthropic and Gemini SDK formats at $0.525 in / $1.75 out per million tokens, prepaid credit, no subscription.
- [Uplink Red Team (Pro)](https://uplinkswitch.ai/models/uplink-red-team-pro): UplinkSwitch's own uncensored build (served without a content filter), served as model id uplink-red-team-pro on the OpenAI, Anthropic and Gemini SDK formats at $1.50 in / $6.00 out per million tokens, prepaid credit, no subscription.
- [Hanami 70B x1](https://uplinkswitch.ai/models/hanami-70b-x1): Sao10K's roleplay build (answers in character instead of refusing), served as model id hanami-70b-x1 on the OpenAI, Anthropic and Gemini SDK formats at $1.7255 in / $1.7255 out per million tokens, prepaid credit, no subscription.
- [Hermes 3 Llama 3.1 70B](https://uplinkswitch.ai/models/hermes-3-llama-3.1-70b): Nous Research's uncensored build (served without a content filter), served as model id hermes-3-llama-3.1-70b on the OpenAI, Anthropic and Gemini SDK formats at $1.632 in / $1.632 out per million tokens, prepaid credit, no subscription.
- [Hermes 4 405B](https://uplinkswitch.ai/models/hermes-4-405b): Nous Research's uncensored build (served without a content filter), served as model id hermes-4-405b on the OpenAI, Anthropic and Gemini SDK formats at $1.20 in / $4.80 out per million tokens, prepaid credit, no subscription.
- [Hermes 4 405B Uncensored](https://uplinkswitch.ai/models/hermes-4-405b-uncensored): Nous Research's uncensored build (served without a content filter), served as model id hermes-4-405b-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.20 in / $4.80 out per million tokens, prepaid credit, no subscription.
- [Inkling Uncensored](https://uplinkswitch.ai/models/inkling-uncensored): Thinking Machines's uncensored build (served without a content filter), served as model id inkling-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $2.13 in / $12.1403 out per million tokens, prepaid credit, no subscription.
- [Jailbreak-Free Assistant](https://uplinkswitch.ai/models/jailbreak-free-assistant): UplinkSwitch's own uncensored build (served without a content filter), served as model id jailbreak-free-assistant on the OpenAI, Anthropic and Gemini SDK formats at $0.75 in / $2.50 out per million tokens, prepaid credit, no subscription.
- [Kimi K2.6 Uncensored](https://uplinkswitch.ai/models/kimi-k2.6-uncensored): Moonshot AI's uncensored build (served without a content filter), served as model id kimi-k2.6-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.17 in / $7.34 out per million tokens, prepaid credit, no subscription.
- [Llama 3 70B Abliterated v3.5](https://uplinkswitch.ai/models/llama-3-70b-abliterated): failspy's abliterated build (the refusal direction removed from the weights), served as model id llama-3-70b-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $2.80 in / $2.80 out per million tokens, prepaid credit, no subscription.
- [Llama 3.3 70B Abliterated](https://uplinkswitch.ai/models/llama-3.3-70b-abliterated): huihui-ai's abliterated build (the refusal direction removed from the weights), served as model id llama-3.3-70b-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $2.80 in / $2.80 out per million tokens, prepaid credit, no subscription.
- [Mag Mell R1 12B](https://uplinkswitch.ai/models/mag-mell-12b-r1): inflatebot's roleplay build (answers in character instead of refusing), served as model id mag-mell-12b-r1 on the OpenAI, Anthropic and Gemini SDK formats at $1.7255 in / $1.7255 out per million tokens, prepaid credit, no subscription.
- [Magidonia 24B v4.3](https://uplinkswitch.ai/models/magidonia-24b-v4.3): TheDrummer's roleplay build (answers in character instead of refusing), served as model id magidonia-24b-v4.3 on the OpenAI, Anthropic and Gemini SDK formats at $0.3535 in / $0.4235 out per million tokens, prepaid credit, no subscription.
- [Magnum Diamond 24B](https://uplinkswitch.ai/models/magnum-diamond-24b): Doctor-Shotgun's roleplay build (answers in character instead of refusing), served as model id magnum-diamond-24b on the OpenAI, Anthropic and Gemini SDK formats at $1.972 in / $1.972 out per million tokens, prepaid credit, no subscription.
- [Magnum v4 72B](https://uplinkswitch.ai/models/magnum-v4-72b): Anthracite's roleplay build (answers in character instead of refusing), served as model id magnum-v4-72b on the OpenAI, Anthropic and Gemini SDK formats at $8.0243 in / $11.9684 out per million tokens, prepaid credit, no subscription.
- [Manta Flash 1.0](https://uplinkswitch.ai/models/manta-flash-1.0): MegaNova AI's roleplay build (answers in character instead of refusing), served as model id manta-flash-1.0 on the OpenAI, Anthropic and Gemini SDK formats at $0.07 in / $0.56 out per million tokens, prepaid credit, no subscription.
- [Manta Mini 1.0](https://uplinkswitch.ai/models/manta-mini-1.0): MegaNova AI's roleplay build (answers in character instead of refusing), served as model id manta-mini-1.0 on the OpenAI, Anthropic and Gemini SDK formats at $0.07 in / $0.56 out per million tokens, prepaid credit, no subscription.
- [Manta Pro 1.0](https://uplinkswitch.ai/models/manta-pro-1.0): MegaNova AI's roleplay build (answers in character instead of refusing), served as model id manta-pro-1.0 on the OpenAI, Anthropic and Gemini SDK formats at $0.21 in / $1.75 out per million tokens, prepaid credit, no subscription.
- [MiMo-V2.5 Uncensored](https://uplinkswitch.ai/models/mimo-v2.5-uncensored): Xiaomi's uncensored build (served without a content filter), served as model id mimo-v2.5-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $0.75 in / $4.00 out per million tokens, prepaid credit, no subscription.
- [MiMo V2.6 Flash Abliterated](https://uplinkswitch.ai/models/mimo-v2.6-flash-abliterated): Xiaomi's abliterated build (the refusal direction removed from the weights), served as model id mimo-v2.6-flash-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $2.00 in / $6.00 out per million tokens, prepaid credit, no subscription.
- [MiMo V2.6 Flash Uncensored](https://uplinkswitch.ai/models/mimo-v2.6-flash-uncensored): Xiaomi's uncensored build (served without a content filter), served as model id mimo-v2.6-flash-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $2.00 in / $6.00 out per million tokens, prepaid credit, no subscription.
- [MiMo V2.6 Flash Uncensored Thinking](https://uplinkswitch.ai/models/mimo-v2.6-flash-uncensored-thinking): Xiaomi's uncensored build (served without a content filter), served as model id mimo-v2.6-flash-uncensored-thinking on the OpenAI, Anthropic and Gemini SDK formats at $1.25 in / $3.75 out per million tokens, prepaid credit, no subscription.
- [MiMo V2.6 Pro Uncensored](https://uplinkswitch.ai/models/mimo-v2.6-pro-uncensored): Xiaomi's uncensored build (served without a content filter), served as model id mimo-v2.6-pro-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.5225 in / $3.045 out per million tokens, prepaid credit, no subscription.
- [MythoMax L2 13B](https://uplinkswitch.ai/models/mythomax-l2-13b): Gryphe's roleplay build (answers in character instead of refusing), served as model id mythomax-l2-13b on the OpenAI, Anthropic and Gemini SDK formats at $0.404 in / $0.404 out per million tokens, prepaid credit, no subscription.
- [NemoMix Unleashed 12B](https://uplinkswitch.ai/models/nemomix-unleashed-12b): MarinaraSpaghetti's roleplay build (answers in character instead of refusing), served as model id nemomix-unleashed-12b on the OpenAI, Anthropic and Gemini SDK formats at $1.7255 in / $1.7255 out per million tokens, prepaid credit, no subscription.
- [NeuralDaredevil 8B Abliterated](https://uplinkswitch.ai/models/neuraldaredevil-8b-abliterated): mlabonne's abliterated build (the refusal direction removed from the weights), served as model id neuraldaredevil-8b-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $1.76 in / $1.76 out per million tokens, prepaid credit, no subscription.
- [No Filter AI](https://uplinkswitch.ai/models/no-filter-ai): UplinkSwitch's own uncensored build (served without a content filter), served as model id no-filter-ai on the OpenAI, Anthropic and Gemini SDK formats at $5.85 in / $7.80 out per million tokens, prepaid credit, no subscription.
- [The Omega Directive 24B v2.0](https://uplinkswitch.ai/models/omega-directive-24b-v2): ReadyArt's roleplay build (answers in character instead of refusing), served as model id omega-directive-24b-v2 on the OpenAI, Anthropic and Gemini SDK formats at $2.00 in / $2.00 out per million tokens, prepaid credit, no subscription.
- [Qwen 2.5 32B Abliterated](https://uplinkswitch.ai/models/qwen2.5-32b-abliterated): huihui-ai's abliterated build (the refusal direction removed from the weights), served as model id qwen2.5-32b-abliterated on the OpenAI, Anthropic and Gemini SDK formats at $1.75 in / $1.75 out per million tokens, prepaid credit, no subscription.
- [Qwen3.5 27B Blossom V6.4 Derestricted](https://uplinkswitch.ai/models/qwen3.5-27b-blossom-v6.4-derestricted): Qwen (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id qwen3.5-27b-blossom-v6.4-derestricted on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Qwen3.5 27B BlueStar v3 Derestricted](https://uplinkswitch.ai/models/qwen3.5-27b-bluestar-v3-derestricted): Qwen (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id qwen3.5-27b-bluestar-v3-derestricted on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled Derestricted](https://uplinkswitch.ai/models/qwen3.5-27b-claude-4.6-opus-distilled-derestricted): Qwen (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id qwen3.5-27b-claude-4.6-opus-distilled-derestricted on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Qwen3.5 27B Claude 4.6 Opus Reasoning Distilled Derestricted Lite](https://uplinkswitch.ai/models/qwen3.5-27b-claude-4.6-opus-distilled-derestricted-lite): Qwen (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id qwen3.5-27b-claude-4.6-opus-distilled-derestricted-lite on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Qwen3.5 27B Queen Derestricted](https://uplinkswitch.ai/models/qwen3.5-27b-queen-derestricted): Qwen (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id qwen3.5-27b-queen-derestricted on the OpenAI, Anthropic and Gemini SDK formats at $1.071 in / $1.071 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Antislop](https://uplinkswitch.ai/models/qwen3.8-27b-antislop): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-antislop on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $2.10 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Cybersecurity](https://uplinkswitch.ai/models/qwen3.8-27b-cybersecurity): Qwen (community tune)'s uncensored build (served without a content filter), served as model id qwen3.8-27b-cybersecurity on the OpenAI, Anthropic and Gemini SDK formats at $0.40 in / $2.40 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B DarkIdol](https://uplinkswitch.ai/models/qwen3.8-27b-darkidol): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-darkidol on the OpenAI, Anthropic and Gemini SDK formats at $0.40 in / $2.40 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Fable](https://uplinkswitch.ai/models/qwen3.8-27b-fable): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-fable on the OpenAI, Anthropic and Gemini SDK formats at $0.875 in / $5.25 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Fimi](https://uplinkswitch.ai/models/qwen3.8-27b-fimi): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-fimi on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $2.10 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Hemingway](https://uplinkswitch.ai/models/qwen3.8-27b-hemingway): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-hemingway on the OpenAI, Anthropic and Gemini SDK formats at $0.875 in / $5.25 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Kimiko](https://uplinkswitch.ai/models/qwen3.8-27b-kimiko): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-kimiko on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $2.10 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Kiwen](https://uplinkswitch.ai/models/qwen3.8-27b-kiwen): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-kiwen on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $2.10 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Novelist](https://uplinkswitch.ai/models/qwen3.8-27b-novelist): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-novelist on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $2.10 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Obliterated](https://uplinkswitch.ai/models/qwen3.8-27b-obliterated): Qwen (community tune)'s abliterated build (the refusal direction removed from the weights), served as model id qwen3.8-27b-obliterated on the OpenAI, Anthropic and Gemini SDK formats at $0.582 in / $7.27 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Queen](https://uplinkswitch.ai/models/qwen3.8-27b-queen): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-queen on the OpenAI, Anthropic and Gemini SDK formats at $0.875 in / $5.25 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Synthia](https://uplinkswitch.ai/models/qwen3.8-27b-synthia): Qwen (community tune)'s roleplay build (answers in character instead of refusing), served as model id qwen3.8-27b-synthia on the OpenAI, Anthropic and Gemini SDK formats at $0.35 in / $2.10 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 27B Uncensored](https://uplinkswitch.ai/models/qwen3.8-27b-uncensored): Qwen (community tune)'s uncensored build (served without a content filter), served as model id qwen3.8-27b-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.25 in / $3.75 out per million tokens, prepaid credit, no subscription.
- [Qwen 3.8 27B Uncensored Fast](https://uplinkswitch.ai/models/qwen3.8-27b-uncensored-fast): Heabsy's uncensored build (served without a content filter), served as model id qwen3.8-27b-uncensored-fast on the OpenAI, Anthropic and Gemini SDK formats at $2.50 in / $7.50 out per million tokens, prepaid credit, no subscription.
- [Qwen 3.8 27B Uncensored Thinking](https://uplinkswitch.ai/models/qwen3.8-27b-uncensored-thinking): Alibaba's uncensored build (served without a content filter), served as model id qwen3.8-27b-uncensored-thinking on the OpenAI, Anthropic and Gemini SDK formats at $2.50 in / $7.50 out per million tokens, prepaid credit, no subscription.
- [Qwen3.8 Max Uncensored](https://uplinkswitch.ai/models/qwen3.8-max-uncensored): Qwen's uncensored build (served without a content filter), served as model id qwen3.8-max-uncensored on the OpenAI, Anthropic and Gemini SDK formats at $1.00 in / $3.00 out per million tokens, prepaid credit, no subscription.
- [ReMM SLERP L2 13B](https://uplinkswitch.ai/models/remm-slerp-l2-13b): Undi95's roleplay build (answers in character instead of refusing), served as model id remm-slerp-l2-13b on the OpenAI, Anthropic and Gemini SDK formats at $2.036 in / $3.78 out per million tokens, prepaid credit, no subscription.
- [Skyfall 36B v2](https://uplinkswitch.ai/models/skyfall-36b-v2): TheDrummer's roleplay build (answers in character instead of refusing), served as model id skyfall-36b-v2 on the OpenAI, Anthropic and Gemini SDK formats at $1.925 in / $2.80 out per million tokens, prepaid credit, no subscription.
- [Starcannon Unleashed 12B v1.0](https://uplinkswitch.ai/models/starcannon-unleashed-12b): VongolaChouko's roleplay build (answers in character instead of refusing), served as model id starcannon-unleashed-12b on the OpenAI, Anthropic and Gemini SDK formats at $1.7255 in / $1.7255 out per million tokens, prepaid credit, no subscription.
- [Stheno 8B v3.2](https://uplinkswitch.ai/models/stheno-8b-v3.2): Sao10K's roleplay build (answers in character instead of refusing), served as model id stheno-8b-v3.2 on the OpenAI, Anthropic and Gemini SDK formats at $0.7035 in / $0.7035 out per million tokens, prepaid credit, no subscription.
- [Uncensored AI Assistant](https://uplinkswitch.ai/models/uncensored-ai-assistant): UplinkSwitch's own uncensored build (served without a content filter), served as model id uncensored-ai-assistant on the OpenAI, Anthropic and Gemini SDK formats at $5.85 in / $7.80 out per million tokens, prepaid credit, no subscription.
- [Uncensored AI Chat](https://uplinkswitch.ai/models/uncensored-ai-chat): UplinkSwitch's own uncensored build (served without a content filter), served as model id uncensored-ai-chat on the OpenAI, Anthropic and Gemini SDK formats at $11.70 in / $15.60 out per million tokens, prepaid credit, no subscription.
- [Uncensored AI Novelist](https://uplinkswitch.ai/models/uncensored-ai-novelist): UplinkSwitch's own uncensored build (served without a content filter), served as model id uncensored-ai-novelist on the OpenAI, Anthropic and Gemini SDK formats at $0.75 in / $3.00 out per million tokens, prepaid credit, no subscription.
- [Uncensored AI Roleplay](https://uplinkswitch.ai/models/uncensored-ai-roleplay): UplinkSwitch's own roleplay build (answers in character instead of refusing), served as model id uncensored-ai-roleplay on the OpenAI, Anthropic and Gemini SDK formats at $0.60 in / $2.34 out per million tokens, prepaid credit, no subscription.
- [Uncensored Chatbot](https://uplinkswitch.ai/models/uncensored-chatbot): UplinkSwitch's own uncensored build (served without a content filter), served as model id uncensored-chatbot on the OpenAI, Anthropic and Gemini SDK formats at $5.85 in / $7.80 out per million tokens, prepaid credit, no subscription.
- [Uncensored Code Assistant](https://uplinkswitch.ai/models/uncensored-code-assistant): UplinkSwitch's own uncensored build (served without a content filter), served as model id uncensored-code-assistant on the OpenAI, Anthropic and Gemini SDK formats at $11.70 in / $15.60 out per million tokens, prepaid credit, no subscription.
- [Uncensored Coding Agent](https://uplinkswitch.ai/models/uncensored-coding-agent): UplinkSwitch's own uncensored build (served without a content filter), served as model id uncensored-coding-agent on the OpenAI, Anthropic and Gemini SDK formats at $0.90 in / $3.00 out per million tokens, prepaid credit, no subscription.
- [Unfiltered AI](https://uplinkswitch.ai/models/unfiltered-ai): UplinkSwitch's own uncensored build (served without a content filter), served as model id unfiltered-ai on the OpenAI, Anthropic and Gemini SDK formats at $5.85 in / $7.80 out per million tokens, prepaid credit, no subscription.
- [Unrestricted AI](https://uplinkswitch.ai/models/unrestricted-ai): UplinkSwitch's own uncensored build (served without a content filter), served as model id unrestricted-ai on the OpenAI, Anthropic and Gemini SDK formats at $11.70 in / $15.60 out per million tokens, prepaid credit, no subscription.
- [UnslopNemo 12B v4.1](https://uplinkswitch.ai/models/unslopnemo-12b-v4.1): TheDrummer's roleplay build (answers in character instead of refusing), served as model id unslopnemo-12b-v4.1 on the OpenAI, Anthropic and Gemini SDK formats at $1.972 in / $1.972 out per million tokens, prepaid credit, no subscription.
- [Veiled Calla 12B](https://uplinkswitch.ai/models/veiled-calla-12b): soob3123's roleplay build (answers in character instead of refusing), served as model id veiled-calla-12b on the OpenAI, Anthropic and Gemini SDK formats at $1.05 in / $1.05 out per million tokens, prepaid credit, no subscription.
- [Wayfarer Large 70B](https://uplinkswitch.ai/models/wayfarer-large-70b): Latitude's roleplay build (answers in character instead of refusing), served as model id wayfarer-large-70b on the OpenAI, Anthropic and Gemini SDK formats at $2.4535 in / $2.4535 out per million tokens, prepaid credit, no subscription.

## Gateway features

- [Gateway features](https://uplinkswitch.ai/features): Agent tools, files and redaction, routing, caching, guardrails and app sign-in, the MCP gateway, endpoints and observability, each with a request and live prices.
- [Server-side web search, MCP and code tools](https://uplinkswitch.ai/features/agent-tools): Web search, web fetch, remote MCP tools and a bash and JavaScript sandbox, run by the gateway inside one request, up to 8 tool rounds.
- [PDF input for any model, PII redaction](https://uplinkswitch.ai/features/files-redaction): Documents read as text for models that cannot take files; per-key secret and PII redaction, restored in the answer.
- [Auto router, fallback models and presets](https://uplinkswitch.ai/features/routing): model "auto" across three tiers, fallback lists of up to 5 models, saved presets called as @preset/name, and routing by price, latency or throughput.
- [LLM response cache and context compression](https://uplinkswitch.ai/features/caching): An opt-in response cache billed at a tenth of the original charge, and middle-out compression for prompts longer than the context window.
- [LLM guardrails and OAuth sign-in for apps](https://uplinkswitch.ai/features/guardrails): Opt-in blocked patterns, prompt-injection checks and tool policies per key or org; OAuth PKCE sign-in that mints a key for your app’s user.
- [MCP gateway: one endpoint for every server](https://uplinkswitch.ai/features/mcp-gateway): Register MCP servers with sealed headers and allowlists; every tool on one /mcp/org endpoint, filtered by the key’s tool policy and logged.
- [Embeddings, rerank, Whisper and Batch API](https://uplinkswitch.ai/features/endpoints): POST /v1/embeddings, /v1/moderations, /v1/rerank, /v1/audio/transcriptions and translations, /v1/images/edits, and the Batch API through /v1/files and /v1/batches.
- [LLM sessions, log export and spend caps](https://uplinkswitch.ai/features/observability): Session timelines from x-session-id, a usage CSV, log export to a webhook, Langfuse or any OTLP backend, per-session caps and a loop guard.

## Guides

- [Claude Code](https://uplinkswitch.ai/guides/claude-code): Two environment variables: ANTHROPIC_BASE_URL and ANTHROPIC_AUTH_TOKEN. Claude Code speaks Anthropic’s format, so it runs the Claude models.
- [Claude Code Router](https://uplinkswitch.ai/guides/claude-code-router): The router puts GLM, DeepSeek and Kimi behind Claude Code’s Anthropic-format calls: one provider on this base URL, tested end to end. The Claude models need no router.
- [Cursor](https://uplinkswitch.ai/guides/cursor): Paste the key as the OpenAI API key and override the OpenAI base URL. Any model on the list works in chat and agent; Tab stays on Cursor’s own models.
- [OpenCode and OpenCode Zen](https://uplinkswitch.ai/guides/opencode): One provider block in opencode.json with the openai-compatible package, and OpenCode Zen’s list-price rates next to ours for the same ids.
- [Cline](https://uplinkswitch.ai/guides/cline): Pick OpenAI Compatible, paste the base URL and key, type a model id and click Verify.
- [Roo Code and Zoo Code](https://uplinkswitch.ai/guides/roo-code): Roo Code shut down on 2026-05-15. Zoo Code, the community fork, keeps the same OpenAI Compatible panel: one model id, native tool calls only.
- [Kilo Code API key](https://uplinkswitch.ai/guides/kilo-code): Kilo’s credits at provider rates plus a 5% fee, or your own key as a custom provider in VS Code or kilo.json. Kilo fetches the model list for you.
- [OpenClaw](https://uplinkswitch.ai/guides/openclaw): One provider block under models.providers in ~/.openclaw/openclaw.json, with api set to openai-completions.
- [Codex CLI](https://uplinkswitch.ai/guides/codex-cli): One model_providers table in ~/.codex/config.toml with the base URL and env_key, wire_api responses. Any model id on the list, Claude included.
- [pi coding agent](https://uplinkswitch.ai/guides/pi): One provider block in ~/.pi/agent/models.json: base URL, key, the model ids you want. Every model on one key, pay as you go.
- [omp](https://uplinkswitch.ai/guides/omp): One provider block in ~/.omp/agent/models.yml: the base URL, a key from the console, the model ids you want.
- [Hermes Agent](https://uplinkswitch.ai/guides/hermes-agent): A providers entry in ~/.hermes/config.yaml, the key in .env, one model id. Every model, pay as you go, no subscription.
- [DeepSeek Harness](https://uplinkswitch.ai/guides/deepseek-harness): One key for every model, Claude included, as a Custom model API provider in cordis.patch.yml. Pay per token from prepaid credit.
- [Aider](https://uplinkswitch.ai/guides/aider): Two environment variables and --model openai/<id>. One key, every model, Claude included, billed per token.
- [OpenHands](https://uplinkswitch.ai/guides/openhands): Advanced settings: Custom Model openai/<id>, the base URL and the key. Any model on the list, Claude included.
- [Continue](https://uplinkswitch.ai/guides/continue): One block in config.yaml: provider openai, apiBase, apiKey, any model id on the list. Pay per token, no subscription.
- [GitHub Copilot Chat BYOK](https://uplinkswitch.ai/guides/github-copilot): A Custom Endpoint in chatLanguageModels.json puts every chat model on the list in Copilot’s model picker, billed per token.
- [SillyTavern](https://uplinkswitch.ai/guides/sillytavern): Chat Completion, Custom (OpenAI-compatible), the base URL and a key. Uncensored and DeepSeek models for roleplay.
- [Janitor AI](https://uplinkswitch.ai/guides/janitor-ai): Provider Proxy, a new configuration with the chat completions URL, your key and a model id.
- [Chub AI](https://uplinkswitch.ai/guides/chub-ai): Secrets, OpenAI, Reverse Proxy: the base URL and your key. Uncensored and roleplay models, and the empty-response fix.
- [Open WebUI](https://uplinkswitch.ai/guides/open-webui): One OpenAI-compatible connection: the base URL with /v1 and a key, Verify Connection, and a Model IDs allowlist for the models you want in the selector.
- [OpenAI, Anthropic and Gemini SDKs](https://uplinkswitch.ai/guides/sdks): Set OPENAI_BASE_URL or ANTHROPIC_BASE_URL and keep your code. Also the Gemini SDK and the Vercel AI SDK.
- [LangChain](https://uplinkswitch.ai/guides/langchain): ChatOpenAI with base_url set, in Python and JavaScript. Tool calls, streaming and structured output on Claude, GPT and DeepSeek.
- [LiteLLM](https://uplinkswitch.ai/guides/litellm): SDK and proxy on one key: model openai/<id> with api_base, or a config.yaml model_list. Every chat model, pay per token.
- [Vercel AI SDK](https://uplinkswitch.ai/guides/vercel-ai-sdk): createOpenAICompatible with the base URL and key, then generateText or streamText on any model id.
- [Strands Agents](https://uplinkswitch.ai/guides/strands-agents): The SDK’s OpenAIModel with client_args for the base URL and key, in Python and TypeScript. Agents with tools on Claude, GPT and DeepSeek.
- [Strix](https://uplinkswitch.ai/guides/strix): STRIX_LLM, LLM_API_KEY and LLM_API_BASE for the pentest agent, with the red-team builds and what one turn costs.
- [Images, video, speech and 3D](https://uplinkswitch.ai/guides/media): OpenAI’s image API for the image models, and one generations endpoint for the rest. Failed generations are not charged.
- [Nano Banana pricing](https://uplinkswitch.ai/guides/nano-banana-pro): Price per image for Pro, 2, 2.1 and Lite next to Google’s, and how to generate and edit with the OpenAI SDK.
- [Seedream API](https://uplinkswitch.ai/guides/seedream-api): How to get a Seedream API key, every version’s price per image, image sets of up to 15 and edits.
- [Seedance API pricing](https://uplinkswitch.ai/guides/seedance-api): Price per second for Seedance 2.5, 2, 2 Fast and 2 Mini at every resolution, what a 15-second clip costs, and the call.
- [Grok Imagine API](https://uplinkswitch.ai/guides/grok-imagine-api): How to get a key, the 6 and 10 second clip prices next to xAI’s per-second rate, modes and aspect ratios, and the call.
- [Kling API](https://uplinkswitch.ai/guides/kling-api): Kling 3.0, 3.0 Turbo and O3 per second, Kling 2.6, 2.5 and 2.1 per clip, next to Kling’s own unit rate; sound, elements, multi-shot and the call.
- [Sora 2 API](https://uplinkswitch.ai/guides/sora-api): The Sora 2 API shut down on September 24, 2026: what it cost per second, Kling 3.0, Veo 3.1 and Seedance 2.5 priced next to it, and a Sora request mapped field by field.
- [ElevenLabs API](https://uplinkswitch.ai/guides/elevenlabs-api): How to get a key, Eleven v3 and Turbo v2.5 per 1,000 characters and Eleven Music per minute, worked per chapter and per track next to ElevenLabs’ rates.
- [Veo 3.1 API pricing](https://uplinkswitch.ai/guides/veo-3-1): Veo 3.1, Fast and Lite per second and per 8-second clip next to Google’s rate, a month of clips, the generations call, and what the Veo 3 API means today.

## Pages

- [UplinkSwitch](https://uplinkswitch.ai/)
- [All chat models and prices](https://uplinkswitch.ai/models): Every chat model with our price per million tokens, the developer’s list price and the SDK formats that call it, from the live catalog.
- [All media models and prices](https://uplinkswitch.ai/models/media): Image, video, speech and 3D models on one key, priced per image, per second or per generation.
- [Uncensored models on one key](https://uplinkswitch.ai/uncensored): Every uncensored, abliterated and roleplay model with its price, plus OpenAI SDK, SillyTavern and Janitor AI settings.
- [Uncensored AI for cybersecurity](https://uplinkswitch.ai/red-teaming): The Red Team and cybersecurity chat builds with their price per million tokens, plus OpenAI-compatible settings for PentestGPT, Strix and HackingBuddyGPT.
- [Claude Code below list price](https://uplinkswitch.ai/claude-code): The live Claude prices next to Anthropic’s list, the two-variable setup and the daily real-model check.
- [Hosted AI agents](https://uplinkswitch.ai/agents): An agent on any catalog model with your own webhook skills, memory, a schedule and signed webhooks, billed per turn at the model price.
- [AI workflow API](https://uplinkswitch.ai/workflows): A chain of model calls as JSON, run with one request: every step’s output and charge, billed as the sum of the steps plus a per-run fee.
- [Free AI API tools](https://uplinkswitch.ai/tools): The cost calculators, bill rewriter, comparisons, price index, daily check, verifier and free chat, on one page.
- [Image and video model comparisons](https://uplinkswitch.ai/compare/media): Media models in pairs: each one’s price per image or per second, the settings each takes and the request.
- [How the daily real-model check works](https://uplinkswitch.ai/proof/method): What the check sends to every model, what each result means and where the results are published.
- [OpenRouter alternatives](https://uplinkswitch.ai/alternatives/openrouter): OpenRouter and its alternatives ranked by the fee each adds over the developer’s list price, dated and sourced.
- [UplinkSwitch vs OpenRouter](https://uplinkswitch.ai/vs/openrouter): OpenRouter’s list price plus its credit fee next to ours, a $100 top-up on both, the model-id map and how to switch.
- [Requesty alternatives](https://uplinkswitch.ai/alternatives/requesty): Requesty’s list plus 5% on one Cline and OpenCode month, next to OpenRouter, LiteLLM, Portkey and a key under list.
- [Portkey alternatives](https://uplinkswitch.ai/alternatives/portkey): After the Palo Alto Networks deal: a month on the $49 plan plus list-price tokens, next to LiteLLM, OpenRouter and a key under list.
- [Hosted LiteLLM alternative](https://uplinkswitch.ai/alternatives/litellm): LiteLLM’s 2026 security incidents dated from its own posts, what a hosted key changes, and one month at list vs under list.
- [OpenRouter vs LiteLLM](https://uplinkswitch.ai/compare/openrouter-vs-litellm): Hosted vs self-hosted: the break-even is 5.5% of your monthly spend against your server bill, worked at three spends.
- [CometAPI alternative](https://uplinkswitch.ai/alternatives/cometapi): CometAPI’s 20% under list checked per model against the developer’s price, one coding month on both, and a daily real-model check.
- [AI/ML API alternative](https://uplinkswitch.ai/alternatives/aimlapi): One Claude Code request at Anthropic’s list, at both prices on AI/ML API’s model pages, and here, dated and linked.
- [kie.ai alternative](https://uplinkswitch.ai/alternatives/kie-ai): Veo 3.1 and Nano Banana Pro per clip and per image: kie.ai’s price next to Google’s list and ours, dated.
- [Venice AI alternative](https://uplinkswitch.ai/alternatives/venice): Venice’s plans and API prices read 2026-10-19 next to our uncensored builds per token, and what each stores.
- [Referral program](https://uplinkswitch.ai/referrals): 10% of a referred account’s top-ups for 12 months, as balance.
- [API status](https://uplinkswitch.ai/status): The API and the Claude, GPT and Gemini routes, from hourly probes.
- [Apps on UplinkSwitch](https://uplinkswitch.ai/apps): Known public tools sending requests, by tokens, last 7 and 30 days; every other app grouped as Other.
- [Changelog](https://uplinkswitch.ai/changelog): What changed, by day. RSS: https://uplinkswitch.ai/changelog/feed.xml
- [Blog](https://uplinkswitch.ai/blog): Prices, token counts and method notes. RSS: https://uplinkswitch.ai/blog/feed.xml
- [Integration guides](https://uplinkswitch.ai/guides): Every tool and SDK set up on one base URL and one key.
- [About](https://uplinkswitch.ai/about): Who runs UplinkSwitch, how prices are set and how to check what you are billed.
- [Contact](https://uplinkswitch.ai/contact): Request a key, report a problem by request id, or ask about a model.
- [Brand](https://uplinkswitch.ai/brand): The mark, wordmark, palette, type and voice, with the files to download.
- [Privacy](https://uplinkswitch.ai/privacy): What is stored, who processes it and for how long.
- [Terms of service](https://uplinkswitch.ai/terms)
- [Acceptable use](https://uplinkswitch.ai/acceptable-use): What the service may not be used for and how abuse is reported.
- [LLM cost calculator](https://uplinkswitch.ai/tools/llm-cost-calculator): One request and a month of them on any chat model: input, cached input and output tokens at the developer’s list price and at ours.
- [Claude Code cost calculator](https://uplinkswitch.ai/tools/claude-code-cost-calculator): The same arithmetic on the Claude models, opened on the size of a real Claude Code request: a long cached prefix, a few thousand new tokens, about a thousand out.
- [AI API bill rewriter](https://uplinkswitch.ai/savings): Drop a usage export from OpenAI or Anthropic and see the same usage priced here. The file is read in your browser.
- [Model comparisons](https://uplinkswitch.ai/compare): Two models side by side: input and output per million tokens, each next to its developer’s list price.
- [LLM API price index](https://uplinkswitch.ai/pricing-index): Every chat model with a developer list price, our price next to it, as a table, JSON or CSV.
- [Daily real-model check](https://uplinkswitch.ai/proof): The same test prompt sent to nearly every model each day, with each result published.
- [Statement and receipt verifier](https://uplinkswitch.ai/statements/verify): Check that a usage statement or a response receipt is genuine and unchanged: an Ed25519 signature, checked in your browser.
- [Free uncensored AI chat](https://uplinkswitch.ai/roleplay): A chat panel on the uncensored builds with SillyTavern character cards, no sign-up, a few free turns a day.
- [Free LLM playground](https://uplinkswitch.ai/playground): Chat with any chat model in the catalog, one at a time, no sign-up: a system prompt, temperature and reply length, the cost under each reply, a few free turns a day.
- [Mara Hale, a SillyTavern character card](https://uplinkswitch.ai/roleplay/mara-lighthouse-keeper): Lighthouse keeper on a storm-bound island, used to being alone. Chat free in the browser, or download the card as Character Card V2 JSON.
- [Vic Adler, a SillyTavern character card](https://uplinkswitch.ai/roleplay/vic-adler-noir-detective): Private detective in a rain-soaked city, 1949, one case from broke. Chat free in the browser, or download the card as Character Card V2 JSON.
- [HALCYON, a SillyTavern character card](https://uplinkswitch.ai/roleplay/halcyon-ship-ai): A long-haul freighter's AI, awake for eleven years, glad of a voice. Chat free in the browser, or download the card as Character Card V2 JSON.
- [Brannoc, a SillyTavern character card](https://uplinkswitch.ai/roleplay/brannoc-tavern-keeper): Keeper of the Salt Crow tavern, where every quest in the city starts. Chat free in the browser, or download the card as Character Card V2 JSON.
- [Ilse Varga, a SillyTavern character card](https://uplinkswitch.ai/roleplay/ilse-varga-vampire-countess): A vampire countess, four centuries old, bored of everyone but you. Chat free in the browser, or download the card as Character Card V2 JSON.
- [Text to image API: Nano Banana, GPT Image, FLUX prices](https://uplinkswitch.ai/models/media/tasks/text-to-image): Every text to image model on one key: Nano Banana, GPT Image, Seedream, FLUX, Kling, Qwen and Wan, priced per image next to the developer's own price.
- [Image editing API: Nano Banana, GPT Image, FLUX prices](https://uplinkswitch.ai/models/media/tasks/image-editing): Edit an image with a prompt: Nano Banana, GPT Image, Seedream, FLUX and Kling edit models on one key, priced per image next to the developer's own price.
- [Background remover, upscaler, extender and resizer API](https://uplinkswitch.ai/models/media/tasks/image-tools): Remove a background, upscale 2x or 4x, extend to a new aspect ratio or resize, on the same key as every image model. Priced per image, prepaid.
- [Text to video API: Veo, Kling, Seedance, Wan prices](https://uplinkswitch.ai/models/media/tasks/text-to-video): Text to video on one key: Veo, Kling, Seedance, Wan, Hailuo and Grok Imagine, priced per second or per generation, with the developer's price where known.
- [Image to video API: Kling, Veo, Seedance, Wan prices](https://uplinkswitch.ai/models/media/tasks/image-to-video): Animate a still image: Kling, Veo, Seedance, Wan and Hailuo image to video models on one key, priced per second or per generation. Prepaid, no subscription.
- [Text to speech API: ElevenLabs v3 and Turbo prices](https://uplinkswitch.ai/models/media/tasks/text-to-speech): ElevenLabs v3 and Turbo 2.5 text to speech on one key, priced per 1,000 characters, plus music priced per minute. Prepaid, no subscription.
- [3D generation API: Tripo text and image to 3D prices](https://uplinkswitch.ai/models/media/tasks/3d-generation): Tripo text to 3D, image to 3D and multiview to 3D on one key, priced per generation. Prepaid, no subscription.
- [Claude API pricing](https://uplinkswitch.ai/pricing/claude): Every Claude model next to Anthropic's list price, per million tokens.
- [GPT API pricing](https://uplinkswitch.ai/pricing/gpt): Every GPT model next to OpenAI's list price, per million tokens.
- [Gemini API pricing](https://uplinkswitch.ai/pricing/gemini): Every Gemini model next to Google's list price, per million tokens.
- [Grok API pricing](https://uplinkswitch.ai/pricing/grok): Every Grok model next to xAI's list price, per million tokens.
- [DeepSeek API pricing](https://uplinkswitch.ai/pricing/deepseek): Every DeepSeek model next to DeepSeek's list price, per million tokens.
- [Kimi API pricing](https://uplinkswitch.ai/pricing/kimi): Every Kimi model next to Moonshot AI's list price, per million tokens.
- [GLM API pricing](https://uplinkswitch.ai/pricing/glm): Every GLM model next to Z.ai's list price, per million tokens.
- [Qwen API pricing](https://uplinkswitch.ai/pricing/qwen): Every Qwen model next to Qwen's list price, per million tokens.
- [MiniMax API pricing](https://uplinkswitch.ai/pricing/minimax): Every MiniMax model next to MiniMax's list price, per million tokens.
- [MiMo API pricing](https://uplinkswitch.ai/pricing/mimo): Every MiMo model next to Xiaomi's list price, per million tokens.
- Every model page and every comparison, one line each: [llms-full.txt](https://uplinkswitch.ai/llms-full.txt)

## Blog

- [Kimi K3.1 Pro API at half of Moonshot's price](https://uplinkswitch.ai/blog/kimi-k3-1-pro-api): Kimi K3.1 Pro API at half of Moonshot's list price, on one key with no subscription. A request and a month, worked.
- [GPT Image 2 API pricing: a flat price per image, every tier](https://uplinkswitch.ai/blog/gpt-image-2-api-pricing): GPT Image 2 API pricing per image: OpenAI's token math turned into a price at every quality and size, next to a flat price per tier on one key.
- [Wan API pricing: Wan 2.2 to 3.0 per second and per clip](https://uplinkswitch.ai/blog/wan-api-pricing): Wan API pricing per second for every Wan model on one key, 2.2 to 3.0 Prime, the cost of a 5, 10 or 15 s clip, next to Alibaba's own list price.
- [FLUX API pricing per image vs Black Forest Labs' list](https://uplinkswitch.ai/blog/flux-api-pricing): FLUX API pricing per image for FLUX.2 pro and flex, Kontext pro and max and FLUX 1.1, next to Black Forest Labs' per-megapixel list, and what edits cost.
- [Uncensored AI chat, no sign-up: the free options compared](https://uplinkswitch.ai/blog/free-uncensored-ai-chat): Uncensored AI chat with no sign-up: SillyTavern cards, 12 free turns a day, every build priced per 100 messages, next to Venice, Character.AI and Janitor.
- [SillyTavern character cards: format, PNG vs JSON, downloads](https://uplinkswitch.ai/blog/sillytavern-character-cards): A SillyTavern character card field by field: PNG vs JSON, {{char}} and {{user}}, how many tokens, where to get cards, how to import them, and five to download.
- [Character AI alternatives with no filter, November 2026](https://uplinkswitch.ai/blog/character-ai-alternative): Six Character AI alternatives with no filter, compared on account, own cards, model choice and price per reply, plus a free roleplay chat with no sign-up.
- [Janitor AI alternatives, free and no proxy: November 2026](https://uplinkswitch.ai/blog/janitor-ai-alternative-free): Five Janitor AI alternatives compared on the proxy step, account, cards, model and price per reply, plus the setting that keeps Janitor and swaps the model.
- [Best LLM for cybersecurity, November 2026](https://uplinkswitch.ai/blog/best-llm-for-cybersecurity): The best LLM for cybersecurity work ranked by refusals, context window and price per million tokens, with the model ids to plug into your tools.
- [WhiteRabbitNeo alternatives by API, November 2026](https://uplinkswitch.ai/blog/whiterabbitneo-alternative): WhiteRabbitNeo is now Deep Hat, with no public API price. The alternatives that answer security questions by API, with per-token prices and a setup block.
- [Strix, HackingBuddyGPT, PentestGPT API setup and cost](https://uplinkswitch.ai/blog/pentest-agents-api-setup): The config that points Strix, HackingBuddyGPT, CAI and PentestGPT at an OpenAI-compatible base URL, which model id to use, and what one engagement costs.
- [OpenRouter pricing: the 5.5% fee, free models, November 2026](https://uplinkswitch.ai/blog/openrouter-pricing): OpenRouter pricing in one table: the 5.5% fee and $0.80 minimum by top-up size, the free-model limits, the BYOK allowance and the same $100 here.
- [New uncensored models: Kimi K3, MiniMax M3, Nemotron, GLM](https://uplinkswitch.ai/blog/new-uncensored-models-october-2026): New uncensored models by API: Kimi K3, MiniMax M3, GLM 5.2, Nemotron 3 Ultra, MiMo and our own ten, priced live with the stock twin beside each build.
- [AI API reseller: how to resell model access on one key](https://uplinkswitch.ai/blog/ai-api-reseller): AI API reseller guide: what OpenAI's and Anthropic's terms allow, three ways to resell on one prepaid key, the margin worked live, and what buyers should check.
- [Uncensored Qwen API: the Qwen 3.8 27B build, tested](https://uplinkswitch.ai/blog/uncensored-qwen): Uncensored Qwen by API: the Qwen 3.8 27B builds ranked by downloads, the one we serve, a dated three-prompt refusal test vs stock Qwen, and live prices.
- [Best uncensored online LLM 2026: every build, one API key](https://uplinkswitch.ai/blog/best-uncensored-llm): The best uncensored online LLMs of 2026, ranked: the abliterated, uncensored and roleplay builds you call by API, with live prices and a daily real-model check.
- [What is an abliterated model? Abliteration explained](https://uplinkswitch.ai/blog/abliterated-models): An abliterated model has the direction that makes it refuse removed from its weights. How abliteration works, what it changes, and how to call one by API.
- [Best AI for roleplay: API models, November 2026](https://uplinkswitch.ai/blog/best-ai-for-roleplay): The best AI for roleplay in SillyTavern, Janitor AI or Chub: uncensored and standard models, how three behaved in SillyTavern, and what 100 messages cost.
- [NSFW LLM API: uncensored models for fiction and roleplay](https://uplinkswitch.ai/blog/nsfw-llm-api): What an NSFW LLM is, how the big chat APIs handle adult fiction, and how to call abliterated and uncensored models from your app or a roleplay frontend.
- [OpenRouter uncensored models: what it lists, November 2026](https://uplinkswitch.ai/blog/openrouter-uncensored): On 2026-10-19, one of OpenRouter's 471 models had Uncensored in its name and none were abliterated. The list, its prices, and abliterated builds on one key.
- [GLM 5.2 jailbreak prompts vs an uncensored GLM-5.3](https://uplinkswitch.ai/blog/glm-jailbreak): A GLM 5.2 jailbreak prompt rides along with every message and can still be refused. What it costs per 100 messages, and the GLM builds that need none.
- [Uncensored Gemma API: Heretic refusal test and live prices](https://uplinkswitch.ai/blog/uncensored-gemma-heretic): Uncensored Gemma by API: what Heretic abliteration changes in Google's open-weight Gemma, a dated refusal test, live prices and the cost of 1,000 chat turns.
- [Uncensored AI image generator API: 5 models and prices](https://uplinkswitch.ai/blog/uncensored-image-generator-api): An uncensored AI image generator API with five models: the live price per image, the size and speed we measured on 2026-10-19, and one working request.
- [Is DeepSeek censored? V4.1 Flash and V4 Pro tested](https://uplinkswitch.ai/blog/is-deepseek-censored): Is DeepSeek censored? Four prompts on 2026-10-19 to stock V4.1 Flash and V4 Pro and to the abliterated builds, refusals quoted, plus how to uncensor DeepSeek.
- [Uncensored LLM for mobile: roleplay on Android and iPhone](https://uplinkswitch.ai/blog/uncensored-llm-mobile): An uncensored LLM for mobile with no GPU: Janitor AI, Chub AI and SillyTavern on Android or iPhone with one API key, every field and the cost per 100 messages.
- [Uncensored GGUF LLM models: VRAM per quant vs API price](https://uplinkswitch.ai/blog/uncensored-gguf-vs-api): Uncensored GGUF LLM models: file size and VRAM per quant for eight builds, dated 2026-10-19, next to what the same builds cost per token on one key.
- [OpenRouter alternative for Janitor AI: no 5.5% fee, 2026](https://uplinkswitch.ai/blog/openrouter-alternative-janitor-ai): An OpenRouter alternative for Janitor AI: no fee on top-ups, the three proxy fields filled in, and abliterated DeepSeek and GLM builds priced per 100 messages.
- [Claude Code alternatives: OpenCode, Cline, Kilo, Zoo Code](https://uplinkswitch.ai/blog/claude-code-alternatives): OpenCode vs Claude Code, then Cline, Kilo Code and Zoo Code: what each tool charges, the same turn priced on five models, and one key that runs all five.
- [No logs LLM API: what is stored, what is not, how to check](https://uplinkswitch.ai/blog/no-logs-llm-api): What a no logs LLM API writes per request, what it never writes, and three checks to run on your own responses: receipt, opt-out header, request id.
- [AI API without KYC: no ID check, an email and prepaid credit](https://uplinkswitch.ai/blog/ai-api-without-kyc): What an AI API without KYC means, what Anthropic's and OpenAI's consoles ask for at signup, and what we ask instead: an email and prepaid credit from $10.
- [Uncensored AI video generator API: price per second](https://uplinkswitch.ai/blog/uncensored-ai-video-api): Uncensored AI video generator API: Wan 3.0 Video Uncensored per second at 480p, 720p and 1080p, live, plus a clip we made on 2026-10-19 and one request.
- [Claude API key: how to get an Anthropic API key](https://uplinkswitch.ai/blog/claude-api-key): Get a Claude API key, also called an Anthropic API key, in five steps, see what a request costs on it, and run the same SDK code on one key for every model.
- [GPT API cost per request, November 2026](https://uplinkswitch.ai/blog/gpt-api-pricing): What one GPT API request costs on GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna: a chat turn and a long document, worked at OpenAI's list price and ours.
- [DeepSeek API cost per request, November 2026](https://uplinkswitch.ai/blog/deepseek-api-pricing): What one DeepSeek API request costs on V4 Pro, V4.1 Flash and V4 Flash: a chat turn and a long document, worked at DeepSeek's list price and ours.
- [Gemini API cost per request, November 2026](https://uplinkswitch.ai/blog/gemini-api-pricing): What one Gemini API request costs on Gemini 4 Argon, 3.1 Pro and 3.8 Flash: a chat turn and a long document, worked at Google's list price and ours.
- [Grok API cost per request, November 2026](https://uplinkswitch.ai/blog/grok-api-pricing): What one Grok API request costs on Grok 4.7, 4.5 and 4.3: a chat turn and a long document, worked at xAI's list price and ours.
- [Kimi API cost per request, November 2026](https://uplinkswitch.ai/blog/kimi-api-pricing): What one Kimi API request costs on Kimi K3, K2.7 Code and K2.5: a chat turn and a long document, worked at Moonshot AI's list price and ours.
- [GLM API cost per request, November 2026](https://uplinkswitch.ai/blog/glm-api-pricing): What one GLM API request costs on GLM 5.3, 5.2 and 5.3 Flash: a chat turn and a long document, worked at Z.ai's list price and ours.
- [Qwen API pricing and cost per request, November 2026](https://uplinkswitch.ai/blog/qwen-api-pricing): What one Qwen API request costs on Qwen3.7 Max, 3.7 Plus and 3.6 27B: a chat turn and a long document, worked at Qwen's list price and ours.
- [Coding plans vs pay-per-token API: Kimi, GLM, Qwen](https://uplinkswitch.ai/blog/coding-plans-vs-api): What the Kimi, Z.ai GLM and Alibaba Qwen coding plans cost and include, whether DeepSeek has a code plan, and what the same monthly money buys per token.
- [Cursor with your own API key: cost, November 2026](https://uplinkswitch.ai/blog/cursor-api-key-cost): What one Cursor agent request costs on your own key, worked at each developer's list price and ours, and the four settings that point Cursor at it.
- [Claude API cost per request, Oct 2026: Opus, Sonnet, Haiku](https://uplinkswitch.ai/blog/claude-api-pricing): What one Claude API request costs on Opus 5.5, Opus 4.8, Sonnet 5.5 and Haiku 5.5: a chat turn and a long document at Anthropic's list price and ours.
- [Free AI API: $1 credit for Claude, GPT and Gemini](https://uplinkswitch.ai/blog/free-ai-api): Every new account can claim $1 of API credit that works on every model. What it buys per model, the rules, and the three steps to a working key.
- [LLM API pricing comparison and calculator, November 2026](https://uplinkswitch.ai/blog/llm-api-pricing-2026): Input, cached and output prices for every chat model next to the developer's list, plus a calculator that prices your workload at list and here. Live catalog.
- [Claude Code pricing and cost: plans vs pay-per-token API](https://uplinkswitch.ai/blog/claude-code-api-cost): Claude Code pricing for Pro, Max, Team and Enterprise, and what the same month of work costs per token at Anthropic's list price and ours.