--- summary: "Use PixelML multi-model API in Clawdbot" read_when: - You want to use PixelML models in Clawdbot - You need PixelML setup guidance --- # PixelML PixelML provides a unified OpenAI-compatible API for accessing multiple AI models including GPT-4o, Claude, and more. ## Features - **Multi-model access**: Use GPT, Claude, and other models through a single API - **OpenAI-compatible**: Standard `/v1` endpoints for easy integration - **Streaming**: Supported on all models - **Vision**: Supported on models with vision capability ## Setup ### 1. Get API Key 1. Sign up at [platform.pixelml.com](https://platform.pixelml.com) 2. Go to your account settings and create an API key 3. Copy your API key ### 2. Configure Clawdbot **Option A: Environment Variable** ```bash export PIXELML_API_KEY="your-api-key" ``` **Option B: Interactive Setup (Recommended)** ```bash clawdbot onboard --auth-choice pixelml-api-key ``` This will: 1. Prompt for your API key (or use existing `PIXELML_API_KEY`) 2. Configure the provider automatically 3. Set the default model **Option C: Non-interactive** ```bash clawdbot onboard --non-interactive \ --auth-choice pixelml-api-key \ --pixelml-api-key "your-api-key" ``` ### 3. Verify Setup ```bash clawdbot chat --model pixelml/gpt-4o-mini "Hello, are you working?" ``` ## Available Models | Model ID | Name | Context | Features | |----------|------|---------|----------| | `gpt-4o-mini` | GPT-4o Mini | 128k | Vision | | `gpt-4o` | GPT-4o | 128k | Vision | | `claude-4.5-haiku` | Claude 4.5 Haiku | 200k | Vision | | `claude-4.5-sonnet` | Claude 4.5 Sonnet | 200k | Vision, Reasoning | ## Model Selection Change your default model anytime: ```bash clawdbot models set pixelml/gpt-4o-mini clawdbot models set pixelml/claude-4.5-sonnet ``` List all available models: ```bash clawdbot models list | grep pixelml ``` ## Usage Examples ```bash # Use GPT-4o Mini (default) clawdbot chat --model pixelml/gpt-4o-mini "Hello" # Use Claude via PixelML clawdbot chat --model pixelml/claude-4.5-sonnet "Write a poem" # Use GPT-4o for vision tasks clawdbot chat --model pixelml/gpt-4o ``` ## Config File Example ```json5 { env: { PIXELML_API_KEY: "your-api-key" }, agents: { defaults: { model: { primary: "pixelml/gpt-4o-mini" } } }, models: { mode: "merge", providers: { pixelml: { baseUrl: "https://ishi.pixelml.com/v1", apiKey: "${PIXELML_API_KEY}", api: "openai-completions", models: [ { id: "gpt-4o-mini", name: "GPT-4o Mini", reasoning: false, input: ["text", "image"], cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 }, contextWindow: 128000, maxTokens: 16384 } ] } } } } ``` ## Troubleshooting ### API key not recognized ```bash echo $PIXELML_API_KEY clawdbot models list | grep pixelml ``` ### Connection issues PixelML API is at `https://ishi.pixelml.com/v1`. Ensure your network allows HTTPS connections. ## Links - [PixelML](https://platform.pixelml.com)