Workers AI runs open models (Llama, Whisper, embeddings, image gen) directly inside Cloudflare Workers — no other vendor, generous free tier, one deploy.
Cloudflare Workers AI
Run open models on Cloudflare's edge, billed per neuron
ai.cloudflare.com · Free 10k neurons/day; usage-based
Capabilities
Capabilities
- LLMs, Whisper STT, embeddings, image gen at the edge; Vectorize vector DB alongside
- 10k neurons/day free; integrates natively with Workers/D1/R2
Best for
- AI features in Cloudflare-stack clones without adding another vendor or key
Not for
- Frontier-quality reasoning (serve Claude/GPT/Gemini for the hard parts)