Llama 3.3 70B Instruct is free on 6 providers today. Permanent free tiers come first, then one-time trials.
| Provider | API model ID | Context | Free limits | Sign-up |
|---|---|---|---|---|
| Cloudflare Workers AI | @cf/meta/llama-3.3-70b-instruct-fp8-fast |
24K | Varies by model | No card |
| OVH AI Endpoints | Meta-Llama-3_3-70B-Instruct |
128K | 2 RPM per IP per model (anonymous, free); 400 RPM with a key (pay-as-you-go) | Free account |
| Vercel AI Gateway | meta/llama-3.3-70b |
128K | Lower per-model limits than the paid tier (429 on excess) | Free account |
| IBM watsonx.ai (Lite plan) | meta-llama/llama-3-3-70b-instruct |
131K | About 2 requests/second (Lite) | Card required |
| Novita AI Trial | meta-llama/llama-3.3-70b-instruct |
131K | 60 RPM | Free account |
| Scaleway Generative APIs Trial | llama-3.3-70b-instruct |
128K | Organization quota in tokens/minute (see console) | Free account |
Free lists change often. Each provider page shows its full limits and code examples, and every change is logged in the changelog. Want to call it from Claude Code or Cursor? Use the setup generator.