> ## Documentation Index
> Fetch the complete documentation index at: https://docs.passy.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Tools & Agents
## Connect Your Tools
Point any coding agent at Passy. You only need your [API key](https://dash.passy.ai) and the right base URL.
**Base URL** → `https://api.passy.ai/v1`
1. Open **Settings → Models**
2. Scroll down and Enable **OpenAI API Key and add your API key**
3. Turn on **Override OpenAI Base URL** → `https://api.passy.ai/v1`
4. Scroll Back up and click **View All models**
5. Scroll down in the list to **Add Custom Model**
6. Then add a model (e.g.`passy/Kimi-K2.6`)
7. You can now select this model and use it
You can use Anthropic models via the same Base URL and OpenAI config in Cursor.
1. Installation
```bash theme={null}
pi install git:github.com/passy-ai/passy-pi-provider
```
2. Add key
```bash theme={null}
export PASSY_API_KEY=your-passy-api-key
```
Add the same line to `~/.zshrc` (or `~/.bashrc`) so it persists across terminal sessions.
3. Use Pi
```bash theme={null}
pi
```
4. Select models
```bash theme={null}
/models
```
our Pi extension auto discovers all our models and adds them to the default list in Pi
Add the same line to `~/.zshrc` (or `~/.bashrc`) so it persists across terminal sessions.
```bash theme={null}
export ANTHROPIC_BASE_URL="https://api.passy.ai"
export ANTHROPIC_AUTH_TOKEN="your-passy-api-key"
export CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1
```
For Claude Code you need to keep the BaseURL `https://api.passy.ai` without the `/v1`
To set a default model use
```bash theme={null}
export ANTHROPIC_MODEL="passy/Kimi-K2.6"
```
To use a specific model before opening a chat
```bash theme={null}
claude --model passy/Kimi-K2.6
```
To swap models during a chat
```bash theme={null}
/model passy/Kimi-K2.6
```
1. Open the **Cline** panel in VS Code → click **⚙️ Settings**
2. Set **API Provider** → **OpenAI Compatible**
3. **Base URL** → `https://api.passy.ai/v1`
4. **API Key** → your Passy API key
5. **Model ID** → e.g. `passy/Kimi-K2.6` → click **Done**
Add to `opencode.json` (project root or `~/.config/opencode/`):
```json theme={null}
{
"$schema": "https://opencode.ai/config.json",
"provider": {
"passy": {
"npm": "@ai-sdk/openai-compatible",
"name": "Passy",
"options": {
"baseURL": "https://api.passy.ai/v1",
"apiKey": "your-passy-api-key"
},
"models": {
"passy/Kimi-K2.6": { "name": "Passy/Kimi K2.6" }
}
}
}
}
```
Restart `OpenCode `and select the Passy provider.
1. Open **Settings** → search `Warp Agent`
2. Inside `Warp Agent` Scroll down to `Custom Inference`
3. **Endpoint URL** → `https://api.passy.ai/v1`
4. **API Key** → your Passy API key
5. **Model** → e.g. `passy/Kimi-K2.6` → **Save**
Your custom model appears in Warp's model picker.
1. Enter `/model` in a new agent window
2. Begin typing your model name e.g `passy/Kimi-K2.6`
3. Select the model and send your requests!
1. Open **Settings** → click on `Providers`
2. Connect `Custom Provider`
3. Fill in the following details\\
* Provider ID = `passy` (lowercase)\\
* Display Name = `Passy` (or custom)\\
* Base URL = `https://api.passy.ai/v1`\\
* API Key = `YOUR_API_KEY`
4. Kilo will discover the models **automatically** and add them for you
5. Scroll to the bottom and Click **Submit**
***
## Support Channels