Rate-limited free · API key · code · Last updated: 29 Oct 2026
5 free models on Cloudflare Workers AI including ~80 free models, @cf/meta/llama-3.3-70b-instruct-fp8-fast, and @cf/deepseek-ai/deepseek-r1-distill-qwen-32b. API key access, up to 128K context, no credit card required. Last updated: 29 Oct 2026.
Neurons are normalized GPU-compute units, shared pool across all models. Resets 00:00 UTC. Hard stop when pool exhausted, errors not overage. No credit card.
Source: https://developers.cloudflare.com/workers-ai/platform/limits/ · Last updated: 29 Oct 2026
| Model | Cost | Context | RPM | TPM | RPD | Day tokens | Verified |
|---|---|---|---|---|---|---|---|
| ~80 free models (Llama 3.x/4, Qwen, Gemma, DeepSeek-R1 distills, FLUX, Whisper, BGE embeddings) | $0 | 8K-128K+ (model-dependent) | 300 text gen (720-1,500 other tasks) | Model-dependent | Not published | 10,000 neurons/day | 2026-08-30 |
| @cf/meta/llama-3.3-70b-instruct-fp8-fast | $0 | 128K | 300 | Model-dependent | Not published | 10,000 neurons/day | 2026-10-29 |
| @cf/deepseek-ai/deepseek-r1-distill-qwen-32b | $0 | 128K | 300 | Model-dependent | Not published | 10,000 neurons/day | 2026-10-29 |
| @cf/qwen/qwen2.5-72b-instruct | $0 | 128K | 300 | Model-dependent | Not published | 10,000 neurons/day | 2026-10-29 |
| @cf/meta/llama-3.1-8b-instruct | $0 | 128K | 300 | Model-dependent | Not published | 10,000 neurons/day | 2026-10-29 |
Machine-readable: /data/cloudflare-workers-ai.json · Markdown