Cloudflare Workers AI

Complete knowledge domain for Cloudflare Workers AI - Run AI models on serverless GPUs across Cloudflare's global network.

---

Auto-Trigger Keywords

Primary Keywords

• workers ai
• cloudflare ai
• ai bindings
• llm workers
• @cf/meta/llama
• workers ai models
• ai inference
• cloudflare llm

Secondary Keywords

• ai streaming
• text generation ai
• ai embeddings
• image generation ai
• workers ai rag
• ai gateway
• llama workers
• flux image generation
• stable diffusion workers
• vision models ai
• ai chat completion

Error-Based Keywords

• AI_ERROR
• rate limit ai
• model not found
• token limit exceeded
• neurons exceeded
• ai quota exceeded
• streaming failed
• model unavailable

Framework Integration Keywords

• workers ai hono
• ai gateway workers
• vercel ai sdk workers
• openai compatible workers
• workers ai vectorize
• ai bindings wrangler
• workers ai streaming

---

What This Skill Does

This skill provides complete Workers AI knowledge including:

• ✅ Text Generation (LLMs) - Llama 3.1/3.2, Qwen, Mistral, DeepSeek (50+ models)
• ✅ Streaming Responses - Real-time text generation without buffering
• ✅ Text Embeddings - BGE models for RAG and semantic search
• ✅ Image Generation - Flux, Stable Diffusion XL for text-to-image
• ✅ Vision Models - Llama 3.2 Vision for image understanding
• ✅ AI Gateway Integration - Caching, logging, and cost tracking
• ✅ OpenAI Compatibility - Drop-in replacement for OpenAI API
• ✅ RAG Patterns - Complete Retrieval Augmented Generation examples
• ✅ Function Calling - Embedded function calling with tools
• ✅ Vercel AI SDK - Integration with workers-ai-provider

---

Known Issues Prevented

---

Quick Example

import { Hono } from 'hono';

type Bindings = {
  AI: Ai;
};

const app = new Hono<{ Bindings: Bindings }>();

// Text generation with streaming
app.post('/chat', async (c) => {
  const { prompt } = await c.req.json<{ prompt: string }>();

  const stream = await c.env.AI.run(
    '@cf/meta/llama-3.1-8b-instruct',
    {
      messages: [{ role: 'user', content: prompt }],
      stream: true,
    }
  );

  return new Response(stream, {
    headers: { 'content-type': 'text/event-stream' },
  });
});

export default app;

wrangler.jsonc:

{
  "ai": {
    "binding": "AI"
  }
}

---

Token Efficiency

Without this skill:

• 15-20 documentation lookups for model selection
• 10+ searches for rate limits and pricing
• 8-12 examples for different task types
• ~8,000 tokens of repetitive context

With this skill:

• 1 skill invocation with complete knowledge
• ~3,200 tokens of focused, battle-tested patterns
• ~60% token savings

---

Coverage

Models Catalog

• Text Generation: 20+ LLMs (Llama, Qwen, Mistral, DeepSeek, Gemma)
• Text Embeddings: BGE-base, BGE-large, BGE-small
• Image Generation: Flux, Stable Diffusion XL, DreamShaper
• Vision Models: Llama 3.2 11B Vision Instruct
• Translation: M2M100, Opus MT
• Classification: DistilBERT, RoBERTa
• Speech Recognition: Whisper models

Patterns

• Chat completions with message history
• RAG (Retrieval Augmented Generation)
• Structured output with Zod schemas
• Function calling with tools
• Image generation and storage
• Vision model image understanding
• AI Gateway caching and logging

Production Topics

• Rate limits by task type
• Neurons-based pricing
• Streaming best practices
• Error handling strategies
• OpenAI API compatibility
• Vercel AI SDK integration
• Cost optimization

---

When to Use This Skill

Use this skill when you see keywords like:

• "integrate Workers AI"
• "add LLM to my Worker"
• "generate embeddings for RAG"
• "create AI image generator"
• "stream AI responses"
• "setup AI Gateway"
• "OpenAI compatible API"
• "model rate limit"
• "Workers AI pricing"

---

References

• Workers AI Documentation
• Models Catalog
• AI Gateway
• Pricing
• Rate Limits