Groq
A free plan with per-model request and token limits.
Free tier
What “free” means here From the provider's documentation
- The free plan has its own per-model limits, listed under “Free Plan Limits” on Groq's rate-limit page.
- The Developer plan raises limits and adds batch and flex processing.
Checked Sep 29, 2026. Limits and terms change; the provider's page always wins.
Use it with cheapoS
- Create a key on Groq's own site. Never paste it into cheapos.lol, a public chat or a repository.
- Add it to your local gateway. In cheapoS, open Models → Open OmniRoute, add the provider and key, then refresh models. Connection guide ↗
- Start small. Pick a tool-capable model and a task you can check before relying on the quota.
Sources
Other providers
- OpenRouterFree tier · One key for many models, including routes with a $0 token rate.
- Google AI Studio (Gemini API)Free tier · A Gemini API free tier for projects, with limits shown in AI Studio.
- Cloudflare Workers AIFree tier · 10,000 Neurons per day at no charge on Free and Paid plans.
- Hugging Face Inference ProvidersMonthly free credits · Small monthly credits to try 200+ models through one token.
- Cerebras InferenceTrial credits · A $5 trial that needs a verified payment method and expires in 30 days.
- Mistral AIPaid, some $0 models · Paid API; a couple of models are listed at $0.