⌁ Guides

Set up Gondola anywhere

Gondola speaks the OpenAI API. Mint one key, set the base URL to api.gondola-ai.com/v1, and any OpenAI-compatible tool can reach the full Venice catalog at marketplace prices.

Claude Code

Repoint ANTHROPIC_BASE_URL and run any model, Claude included.

Open guide →

Cursor

Add Gondola as a custom OpenAI model in the AI editor.

Open guide →

Cline

Point the VS Code agent at Gondola with a per-model config.

Open guide →

Aider

Two env vars and the openai/ prefix in your terminal.

Open guide →

OpenCode

Add Gondola as a provider in opencode.json.

Open guide →

Continue

Chat, edit, and autocomplete from config.yaml.

Open guide →

SillyTavern

Connect via the Custom OpenAI-compatible source.

Install the provider plugin or add two blocks of config.yaml.

Add Gondola as an upstream channel in your self-hosted gateway.

Two env vars, or a named provider in catalog.json.

Open guide →

Zot

Built-in provider: one env var and you are running.

Open guide →

How Gondola works

The Venice relationship and the request pipeline: routing, fees, sanitization, what is stored.

Open guide →

Structured judgment

Classify, route and score text with typed probabilities instead of parsing a chat reply.

Open guide →

Any other OpenAI-compatible tool

If your tool lets you set an OpenAI base URL and API key, it works with Gondola. Mint a key at /keys (keys start with gnd_), point the base URL at the gateway, and pass the key as a normal OpenAI key.

shell

# Most OpenAI clients read these two variables:
OPENAI_BASE_URL=https://api.gondola-ai.com/v1
OPENAI_API_KEY=gnd_xxxx

One rule for the base URL

Use api.gondola-ai.com/v1 and stop there. Do not append /chat/completions; OpenAI clients add the path themselves, so a doubled path is the most common cause of 404s.

Discover exact model ids with curl https://api.gondola-ai.com/v1/models. Top up USDC on Base at /wallet and watch live prices on the marketplace.

Loading page