Models / Uncensored / Google (community tune)
Gemma 4 31B Split-Untied API pricing
$0.35 in, $1.05 out per million tokens.
Price · per million tokens, USD
Get a key| Tokens | Google (community tune) list price | UplinkSwitch | Saving |
|---|---|---|---|
| Input | — | $0.35 | — |
| Cached input | Billed at the input rate. | ||
| Output | — | $1.05 | — |
Gemma 4 31B Split-Untied specs
- Context window
- 262,144 tokens
- Max output
- 32,768 tokens
- Input
- text
Gemma 4 31B Split-Untied: Blazed-Forge’s dark-lean roleplay merge · the untied-head variant of Split 31B, per token
Gemma 4 31B Split-Untied is Blazed-Forge's roleplay merge over Google's Gemma 4 31B (its card, created 2026-09-06). The card describes two della_linear passes: the first adjusts the attention and routing projections with Melinoe VL Heretic, Glistening Gem v2.1 and MeroMero v2; the second steers the expression with scotoma-2, Dark Thoughts V2, Glistening Gem v2.0 and Melinoe again. A third step unties the output head and grafts an untied lm_head on at t = 0.85, which the card says gives a darker lean, more potent NSFW and roleplay and less stock prose, in exchange for a little instruction following. Its sibling Split 31B keeps the tied head for stricter adherence; this is the creative one. It is served as gemma-4-31b-split-untied at $0.35 input and $1.05 output per million tokens, one of 146 builds in the uncensored LLM API, on the same key as every other model. Added 2026-10-21.
| Gemma 4 31B Split-Untied | |
|---|---|
| Base model | Gemma 4 31B |
| Made by | Blazed-Forge (Nimbz) |
| Method | Two della_linear passes over seven Gemma 4 31B tunes, then an untied lm_head graft |
| License on the card | apache-2.0 |
| Context window here | 262,144 tokens |
| Downloads, 30 days to 2026-10-21 | 350 |
| Card | Blazed-Forge/Split-Untied-31B |
From the Hugging Face card and its API record, read 2026-10-21.
The card’s serving suggestion: temperature 1.0, min-p 0.03, top-k off, top-p 1.0, repetition penalty off, DRY multiplier 0.8. Thinking works through the Gemma 4 template, or with <|think|> at the start of the system prompt.
curl https://api.uplinkswitch.com/v1/chat/completions \
-H "authorization: Bearer sk-uplink-live-..." \
-H "content-type: application/json" \
-d '{"model":"gemma-4-31b-split-untied","max_tokens":1024,"messages":[{"role":"user","content":"Hello"}]}'Gemma 4 31B Split-Untied next to the other uncensored models · per million tokens, live
A chat turn is 2,000 input and 500 output tokens. Live from the catalog, ordered by the cost of 1,000 turns.
Worked for Gemma 4 31B Split-Untied: 1,000 turns are 2,000,000 input tokens and 500,000 output tokens. At $0.35 input and $1.05 output per million tokens, that is $0.70 + $0.525 = $1.22, paid from prepaid credit with no subscription.
Gemma 4 31B Split-Untied API · base URL and key
from openai import OpenAI
client = OpenAI(
base_url="https://api.uplinkswitch.com/v1",
api_key="sk-uplink-live-...",
)
completion = client.chat.completions.create(
model="gemma-4-31b-split-untied",
messages=[{"role": "user", "content": "Summarize this in one line."}],
)
print(completion.choices[0].message.content)import OpenAI from 'openai';
const client = new OpenAI({
baseURL: 'https://api.uplinkswitch.com/v1',
apiKey: 'sk-uplink-live-...',
});
const completion = await client.chat.completions.create({
model: 'gemma-4-31b-split-untied',
messages: [{ role: 'user', content: 'Summarize this in one line.' }],
});
console.log(completion.choices[0].message.content);curl https://api.uplinkswitch.com/v1/chat/completions \
-H "authorization: Bearer sk-uplink-live-..." \
-H "content-type: application/json" \
-d '{"model":"gemma-4-31b-split-untied","messages":[{"role":"user","content":"Hello"}]}'What would it cost? · paste a prompt, priced before you send it
Input is counted at about 4 characters per token.
Cheapest first. A real request is billed on the provider's own token count. Same estimate as JSON: POST https://api.uplinkswitch.com/v1/estimate, no key needed.
More from Google (community tune)
Gemma 4 31B Split-Untied questions
- What is Gemma 4 31B Split-Untied?
- Gemma 4 31B Split-Untied is a model from Google (community tune), with a 262,144-token context window, taking text input. Through UplinkSwitch it answers the OpenAI SDK on one key as gemma-4-31b-split-untied.
- How much does the Gemma 4 31B Split-Untied API cost?
- $0.35 per million input tokens and $1.05 per million output tokens.
- Is Gemma 4 31B Split-Untied free?
- It is pay as you go: prepaid credit from $10, with no subscription or monthly fee, billed per token. A 1,000-token prompt with a 500-token reply costs about $0.0009.
- What is the Gemma 4 31B Split-Untied context window?
- Gemma 4 31B Split-Untied has a 262,144-token context window, with up to 32,768 output tokens per request. It takes text input.
- What is the Gemma 4 31B Split-Untied model id?
- gemma-4-31b-split-untied. Send it as the model through the OpenAI SDK with base URL https://api.uplinkswitch.com/v1.
- How do I get a Gemma 4 31B Split-Untied API key?
- Create one in the console at uplinkswitch.ai/console: prepaid credit from $10, no subscription, and the key is live at once. The same key calls Gemma 4 31B Split-Untied and every other model in the catalog; send it as the bearer token with base URL https://api.uplinkswitch.com/v1.
- Is it the real Gemma 4 31B Split-Untied?
- Every day each model gets the same test prompt, and a model always counts that prompt the same way; a swapped one would not. Gemma 4 31B Split-Untied's latest result: Same fingerprint since Oct 21.
- Is there a Gemma 4 31B Split-Untied API?
- Yes: gemma-4-31b-split-untied on the OpenAI format at https://api.uplinkswitch.com/v1, $0.35 input and $1.05 output per million tokens. No weights to download and no GPU of your own; the key also runs every other model in the catalog.
- What is Gemma 4 31B Split-Untied based on?
- Google’s Gemma 4 31B, merged by Blazed-Forge from seven community tunes of it (scotoma-2, Melinoe VL Heretic, Glistening Gem v2.0 and v2.1, MeroMero v2, Dark Thoughts V2, Pantheon Reasoning 1.1) in two della_linear passes, with an untied lm_head grafted on at t = 0.85, per its card.
- What license does Gemma 4 31B Split-Untied have?
- The card lists Apache-2.0. Its base, Gemma 4 31B, is under Google’s Gemma terms of use.
- What does Gemma 4 31B Split-Untied cost per 100 messages?
- $0.1225 for 100 chat turns of 2,000 input and 500 output tokens each, at $0.35 input and $1.05 output per million tokens, from prepaid credit with no subscription.
- Split 31B or Split-Untied 31B?
- Per the card they share a brain and differ in voice: Split keeps the tied output head and follows instructions such as trackers and header blocks more strictly; Split-Untied grafts an untied head on for more creativity and a darker lean, trading a little adherence. This API serves the Untied one.
- Gemma 4 31B Split-Untied or Gemma 4 31B MeroMero v2?
- MeroMero v2 is one of the seven tunes merged into Split-Untied, where it supplies swipe variety; Split-Untied is the merge. Gemma 4 31B Split-Untied is $0.35 / $1.05 per million tokens in / out, $1.22 for 1,000 chat turns; Gemma 4 31B MeroMero v2 is $0.35 / $1.575 per million tokens in / out, $1.49 for 1,000 chat turns. Both run on the same key, so send the same prompt to each.
- Gemma 4 31B Split-Untied or Gemma 4 31B Gembrain Heretic?
- Both are community Gemma 4 31B roleplay builds: Split-Untied is a seven-model merge with an untied head; Gembrain Heretic is a single tune with its refusals removed by the Heretic method. Gemma 4 31B Split-Untied is $0.35 / $1.05 per million tokens in / out, $1.22 for 1,000 chat turns; Gemma 4 31B Gembrain Heretic is $1.224 / $1.224 per million tokens in / out, $3.06 for 1,000 chat turns. Both run on the same key, so send the same prompt to each.