openclaw/docs/providers/pollinations.md
2026-01-29 17:34:57 +00:00

91 lines
2.8 KiB
Markdown

---
summary: "Configure Pollinations.ai (Free/Open Source API)"
---
# Pollinations.ai
Pollinations.ai provides AI API access with OpenAI-compatible endpoints. It supports various models including OpenAI's GPT series, Qwen, and Mistral.
## Usage
You can configure Pollinations interactively during onboarding:
```bash
moltbot onboard --auth-choice pollinations
```
Or configure it manually in your `moltbot.json` config file.
## Configuration
Pollinations uses a standard OpenAI-compatible endpoint. While it may offer a limited free tier, obtaining an API key is recommended for reliable access.
### Config snippet
```javascript
// moltbot.json (partial)
{
env: {
// Set your Pollinations API key here (sk-...)
POLLINATIONS_API_KEY: "sk-..."
},
agents: {
defaults: {
model: { primary: "pollinations/openai" },
// Optional: alias for easier switching
models: {
"pollinations/openai": { alias: "Pollinations GPT-5 Mini" }
}
}
},
models: {
providers: {
pollinations: {
baseUrl: "https://gen.pollinations.ai/v1",
api: "openai-completions",
apiKey: "${POLLINATIONS_API_KEY}",
models: [
{
id: "openai",
name: "OpenAI GPT-5 Mini",
contextWindow: 128000,
maxTokens: 8192
},
// ... other models
]
}
}
}
}
```
## Available Models
Pollinations offers a wide range of models. Common ones include:
- `pollinations/openai`: **GPT-5 Mini** (0.15/0.60 Pollen/M)
- `pollinations/openai-large`: **GPT-5.2** (1.75/14.0 Pollen/M)
- `pollinations/openai-fast`: **GPT-5 Nano** (0.06/0.44 Pollen/M)
- `pollinations/claude`: **Claude Sonnet 4.5** (3.0/15.0 Pollen/M)
- `pollinations/claude-large`: **Claude Opus 4.5** (5.0/25.0 Pollen/M)
- `pollinations/claude-fast`: **Claude Haiku 4.5** (1.0/5.0 Pollen/M)
- `pollinations/gemini`: **Gemini 3 Flash** (0.5/3.0 Pollen/M)
- `pollinations/gemini-large`: **Gemini 3 Pro** (2.0/12.0 Pollen/M)
- `pollinations/gemini-fast`: **Gemini 2.5 Flash Lite** (0.1/0.4 Pollen/M)
- `pollinations/qwen-coder`: **Qwen3 Coder 30B** (0.06/0.22 Pollen/M)
- `pollinations/mistral`: **Mistral Small 3.2 24B** (0.15/0.35 Pollen/M)
- `pollinations/deepseek`: **DeepSeek V3.2** (0.57/1.68 Pollen/M)
- `pollinations/grok`: **xAI Grok 4 Fast** (0.2/0.5 Pollen/M)
- `pollinations/perplexity-fast`: **Perplexity Sonar** (1.0/1.0 Pollen/M)
- `pollinations/nova-fast`: **Amazon Nova Micro** (0.04/0.15 Pollen/M)
*Costs are per million tokens (Input/Output). Prices subject to change.*
## Notes
- **API Key**: An API key is required. Get one at [enter.pollinations.ai](https://enter.pollinations.ai/).
- **Cost**: Usage consumes "Pollen" credits, each pollen is equal to 1$ (Temporarily for beta 1 pollen is 0.5$).
- **Privacy**: See [pollinations.ai](https://pollinations.ai) for privacy details.