Sao10K: Llama 3.3 Euryale 70B

Euryale is a Llama-based finetune from Sao10K aimed at open-ended roleplay and writing, with alignment loosened rather than surgically removed. Served through Redline on one OpenAI-compatible endpoint, priced at $0.748 in and $0.863 out per million tokens.

Model idsao10k/l3.3-euryale-70b
Taguncensored
Context131k tokens
Input price$0.748 / million tokens
Output price$0.863 / million tokens
Tool callingNot supported

What the uncensored tag actually means

sao10k/l3.3-euryale-70b is tagged uncensored, which here means a loosely aligned finetune. It refuses far less than a frontier instruct model, but it is not abliterated: the refusal behaviour was reduced by training, not surgically removed, so it can still decline. Nothing about the tag promises it will answer anything.

Call it with curl

curl https://redline-gateway.pages.dev/v1/chat/completions \
  -H "Authorization: Bearer rl-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "sao10k/l3.3-euryale-70b",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Call it with the OpenAI SDK

# pip install openai
from openai import OpenAI

client = OpenAI(
    base_url="https://redline-gateway.pages.dev/v1",
    api_key="rl-...",
)

resp = client.chat.completions.create(
    model="sao10k/l3.3-euryale-70b",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

In Claude Code

No tool calling This model does not advertise tool calling, so it cannot drive an agent like Claude Code that needs to edit files or run commands. A tool request to it returns a 400 that names the model and costs nothing. Use it for chat and generation through the OpenAI-compatible endpoint below.

FAQ

What does the uncensored tag mean?

Uncensored here means a loosely aligned finetune that refuses much less than a frontier model, but can still decline. It is not the same as abliterated.

Does Sao10K: Llama 3.3 Euryale 70B store my prompts?

No. Redline stores the model id, token counts and cost, not the prompt or completion, and returns a signed receipt of the prompt hash.

How much does it cost?

$0.748 per million input tokens and $0.863 per million output tokens, margin already included.

Other uncensored models