Wild West API

Use Aider with an uncensored LLM API

Aider is a terminal pair programmer that edits files in your git repo. It reaches OpenAI-compatible servers through LiteLLM, so the setup is two environment variables and a model name with the openai/ prefix.

Works with Wild West API.

Set it up

  1. Sign in at wildwestapi.com/signup/ with the six-digit code sent to your email, open the dashboard and create an API key. It starts with sk-ww- and is shown in full once, so copy it somewhere safe.
  2. Install Aider with python -m pip install aider-install, then run aider-install.
  3. Set OPENAI_API_BASE to https://wildwestapi.com/v1 and OPENAI_API_KEY to your sk-ww- key.
  4. On Windows, use setx for both and open a new terminal, because setx does not change the current one.
  5. From your project folder, run aider --model openai/outlaw-1.

The commands

Aider's OpenAI-compatible guide says to set OPENAI_API_BASE and OPENAI_API_KEY, then prefix the model name with openai/.

# macOS / Linux
export OPENAI_API_BASE=https://wildwestapi.com/v1
export OPENAI_API_KEY="$WILDWEST_API_KEY"

cd /path/to/your/project
aider --model openai/outlaw-1
:: Windows (new terminal afterwards)
setx OPENAI_API_BASE https://wildwestapi.com/v1
setx OPENAI_API_KEY sk-ww-your-key-here

Aider also reads a .env file from your home directory, the git root or the current directory, with later ones winning, which is a tidier place for the key than your shell profile. Add .env to .gitignore.

The unknown model warning

Aider does not know openai/outlaw-1, so it prints Unknown context window size and costs, using sane defaults. Per its warnings page this assumes an unlimited window and a free model, and it is usually harmless. To give it real numbers, add a .aider.model.metadata.json next to your project. Entries use the full provider-prefixed name.

{
  "openai/outlaw-1": {
    "max_input_tokens": 500000,
    "litellm_provider": "openai",
    "mode": "chat"
  }
}

Set max_input_tokens to the real window of the model you use. It is a limit on what Aider will send, not a billing setting.

Cost behaviour

Aider sends a map of your repo and the files in the chat on every turn, so a long session on a big project sends a lot of input tokens. With pay-as-you-go billing that is fine, but watch your balance, and use /drop to remove files you are done with.

One limit to know

Aider's cost display will not match your Wild West spend, because it thinks the model is free. Use the dashboard for the real figure.

FAQ

Why the openai/ prefix?

Aider hands the name to LiteLLM, which uses the prefix to choose an OpenAI-compatible route instead of guessing a provider. The prefix is a routing label, not part of the model id the server sees.

Can I run the uncensored model for writing, not code?

Aider is built around editing code in a git repo, so it is a poor fit for prose. Use a chat frontend for fiction.

Related

Uncensored AI models on one key

OpenAI and Anthropic compatible, pay as you go. New to it? Start with uncensored AI, explained.