Models

Llama 3.2 3B Uncensored

A GGUF chat model on Autarcho. You pay for the tokens it generates, at the price set for this model.

Model details

Prices can change without a deploy. You can see the current price in the app before you send anything.

Model id
llama-32-3b-uncensored@1.0 or the full hash 1a01383341be9bf04a904e168089401c3f0533b3af4089d3536d8b36e497ccd6
Format
GGUF, 2.1 GB. Checked before it was listed.
Price
3.00 USDC per 1M generated tokens. Input tokens are not billed.
Availability
Answered by an operator who has this model loaded. The app shows what is being served when you send.

Example request

The endpoint is OpenAI-compatible, so the model id is the only thing that changes.

curl -s https://api.autarcho.com/v1/chat/completions \
  -H "Authorization: Bearer ak_..." \
  -H "content-type: application/json" \
  -d '{
    "model": "llama-32-3b-uncensored@1.0",
    "messages": [{ "role": "user", "content": "Hello" }],
    "stream": true
  }'

A key is ak_ followed by 64 hex characters, made on your own machine. The start page covers the key, the credit and this request from a shell. Models are billed per generated token only, so a long document costs what the answer costs.

Serving this model

An operator who runs this model keeps 80% of what each job it serves is charged. The network owns no hardware; a worker is somebody's own machine, making outbound connections only.

Prices and billing