ai-sdk-provider-poe

Poe provider for the Vercel AI SDK.

Access Anthropic, OpenAI, Google, xAI, and 50+ other AI models through a single Poe API key.

Setup

npm install ai-sdk-provider-poe ai

export POE_API_KEY="your-poe-api-key"

Get an API key from poe.com/api_key.

Usage

Basic text generation

import { poe } from 'ai-sdk-provider-poe';
import { generateText } from 'ai';

const { text } = await generateText({
  model: poe('anthropic/claude-sonnet-4'),
  prompt: 'Hello!',
});

Reasoning with Anthropic (Claude Opus 4.6)

Anthropic models use reasoningBudgetTokens to control the thinking budget.

const { text, reasoning } = await generateText({
  model: poe('anthropic/claude-opus-4.6'),
  prompt: 'Solve this step by step: what is 17 * 19?',
  providerOptions: {
    poe: { reasoningBudgetTokens: 5000 },
  },
});

Reasoning with OpenAI (GPT-5.4)

OpenAI models use reasoningEffort and reasoningSummary.

const { text, reasoning } = await generateText({
  model: poe('openai/gpt-5.4'),
  prompt: 'Solve this step by step: what is 17 * 19?',
  providerOptions: {
    poe: { reasoningEffort: 'high', reasoningSummary: 'auto' },
  },
});

Reasoning with Google (Gemini 3.1 Pro)

const { text } = await generateText({
  model: poe('google/gemini-3.1-pro'),
  prompt: 'Solve this step by step: what is 17 * 19?',
  providerOptions: {
    poe: { reasoningEffort: 'high' },
  },
});

Reasoning with xAI (Grok 4)

const { text } = await generateText({
  model: poe('grok-4-fast-reasoning'),
  prompt: 'Solve this step by step: what is 17 * 19?',
  providerOptions: {
    poe: { reasoningEffort: 'high' },
  },
});

Tool calling

import { generateText, tool } from 'ai';
import { z } from 'zod';

const { text } = await generateText({
  model: poe('openai/gpt-5.2'),
  prompt: "What's the weather in San Francisco?",
  tools: {
    getWeather: tool({
      description: 'Get the weather for a location',
      parameters: z.object({ location: z.string() }),
      execute: async ({ location }) => `72°F and sunny in ${location}`,
    }),
  },
});

Provider Options

providerOptions.poe routes reasoning settings to the correct backend automatically:

| Option | Backend mapping | Use with | |---|---|---| | reasoningBudgetTokens | anthropic.thinking.budgetTokens | Claude models | | reasoningEffort | openai.reasoningEffort | OpenAI, Google, xAI | | reasoningSummary | openai.reasoningSummary | OpenAI /v1/responses models |

Model Routing

Routing is resolved automatically from each model's supported_endpoints. The model catalog is bundled and refreshed at runtime via /v1/models.

For the full model list, visit poe.com/api/models.

Configuration

import { createPoe } from 'ai-sdk-provider-poe';

const poe = createPoe({
  apiKey: 'your-poe-api-key',           // default: POE_API_KEY env var
  baseURL: 'https://api.poe.com/v1',    // default
});

License

MIT