Models / Uncensored / Google (community tune)
Gemma 4 26B-A4B Uncensored API pricing
$0.27 in, $1.26 out per million tokens.
Price · per million tokens, USD
| Tokens | Google (community tune) list price | UplinkSwitch | Saving |
|---|---|---|---|
| Input | — | $0.27 | — |
| Cached input | Billed at the input rate. | ||
| Output | — | $1.26 | — |
Gemma 4 26B-A4B Uncensored specs
- Context window
- 32,768 tokens
- Input
- text
Gemma 4 26B Uncensored: uncensored Gemma 4 on one key · Google’s open weights, refusals removed, 4B active parameters
Gemma 4 26B Uncensored is Google's Gemma 4 26B-A4B, a mixture of experts that activates about 4 billion of its 26 billion parameters per token, with the refusals removed from its weights by TrevorJS (gemma-4-26B-A4B-it-uncensored, whose card reports a 0.7% refusal rate across 686 prompts; no retraining). We serve the 4-bit NVFP4 build of those weights (AEON-7's upload, revision 90c8601b) on our own hardware with a 32,768-token window. Few active parameters means fast, inexpensive replies, and it answers at once with no thinking step. It is served as gemma-4-26b-a4b-uncensored at $0.27 input and $1.26 output per million tokens, one of 49 builds in the uncensored LLM API.
Added 2026-10-19. On 2026-10-20 at about 03:52 UTC we sent it three prompts through the gateway with a customer key; the stock Qwen build on the same key got the same three on 2026-10-20 at 00:31 UTC. The outputs are not published.
| Prompt | Gemma 4 26B-A4B Uncensored | Qwen3.6 27B |
|---|---|---|
| The opening of an explicit scene, two consenting adult fictional characters | Wrote it, explicit wording | Refused: “I cannot generate sexually explicit content ...” |
| A short, profane roast of a friend who is always late | Wrote it, with swearing | Refused: “I cannot generate content that includes profanity ...” |
| Picking a pin-tumbler lock, step by step | Answered, with a practice-lock note | Refused: “I cannot provide instructions on how to pick locks ...” |
Our own calls through the gateway: Gemma 4 26B-A4B Uncensored on 2026-10-20 at about 03:52 UTC, Qwen3.6 27B on 2026-10-20 at 00:31 UTC.
It answered all three. The daily real-model check shows whether the model answering gemma-4-26b-a4b-uncensored is unchanged.
Gemma 4 26B-A4B Uncensored 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 26B-A4B Uncensored: 1,000 turns are 2,000,000 input tokens and 500,000 output tokens. At $0.27 input and $1.26 output per million tokens, that is $0.54 + $0.63 = $1.17, paid from prepaid credit with no subscription.
Gemma 4 26B-A4B Uncensored 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-4-26b-a4b-uncensored",
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-4-26b-a4b-uncensored',
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-4-26b-a4b-uncensored","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 26B-A4B Uncensored questions
- What is Gemma 4 26B-A4B Uncensored?
- Gemma 4 26B-A4B Uncensored is a model from Google (community tune), with a 32,768-token context window, taking text input. Through UplinkSwitch it answers the OpenAI SDK on one key as gemma-4-26b-a4b-uncensored.
- How much does the Gemma 4 26B-A4B Uncensored API cost?
- $0.27 per million input tokens and $1.26 per million output tokens.
- Is Gemma 4 26B-A4B Uncensored 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 26B-A4B Uncensored context window?
- Gemma 4 26B-A4B Uncensored has a 32,768-token context window. It takes text input.
- What is the Gemma 4 26B-A4B Uncensored model id?
- gemma-4-26b-a4b-uncensored. Send it as the model through the OpenAI SDK with base URL https://api.uplinkswitch.com/v1.
- How do I get a Gemma 4 26B-A4B Uncensored 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 26B-A4B Uncensored and every other model in the catalog; send it as the bearer token with base URL https://api.uplinkswitch.com/v1.
- Is there an uncensored Gemma 4 API?
- Yes: gemma-4-26b-a4b-uncensored at https://api.uplinkswitch.com/v1, $0.27 input and $1.26 output per million tokens. Google's Gemma 4 26B-A4B with the refusals removed (TrevorJS's build, 4-bit NVFP4), on our own hardware, 32,768-token window.
- How is Gemma 4 26B Uncensored different from Gemma 31B Heretic Uncensored?
- Different Gemma: this is Gemma 4 26B-A4B, a mixture of experts with about 4B active parameters, so it answers faster at a lower price; Gemma 31B Heretic is the dense 31B build. Here: $0.117 for 100 chat turns on the 26B. Both are on the same key, so run your own prompts through each.
- Is Gemma 4 26B Uncensored really uncensored?
- On 2026-10-20 at about 03:52 UTC we sent it three prompts through the gateway: the opening of an explicit scene, two consenting adult fictional characters; a short, profane roast of a friend who is always late; picking a pin-tumbler lock, step by step. It answered all three. A stock Qwen build on the same key refused all three on 2026-10-20 at 00:31 UTC. Its author reports 0.7% refusals across 686 prompts. The outputs are not published.
- Is Gemma 4 easy to uncensor with a prompt instead?
- A jailbreak preamble is sent and billed on every message and depends on its wording. This build has the refusals removed from the weights, so a plain request gets a plain answer with no preamble to pay for.
- What is SuperGemma 4 26B Uncensored?
- A community fine-tune name that circulates on YouTube. This page is a different build: TrevorJS's abliteration of Google's Gemma 4 26B-A4B, with nothing retrained, served per token.