Frequently asked questions
Everything about using FreeLLMAPI as a free LLM API — the short version.
FAQ
Is FreeLLMAPI really free?
Yes. The router is open source and self-hosted, and it only calls providers' free tiers, so inference costs you nothing. Premium ($19/yr) is optional and only keeps the model catalog updated in real time; the routing itself is free forever.
Do I need a credit card?
No. Every provider in the catalog offers a free tier that works without a card. You add each provider's free key once, and FreeLLMAPI routes across them.
Which models can I use for free?
235+ model families across 28 providers, including Kimi K2, Qwen3 Coder, DeepSeek V3/R1, Gemini Flash, Llama, GLM, Nemotron and GPT-OSS. Browse the full free model catalog.
Is it OpenAI-compatible?
Yes. It exposes a standard /v1 chat-completions endpoint, so any OpenAI SDK or tool works by changing the base URL and key.
How is this different from OpenRouter?
FreeLLMAPI focuses only on genuinely free tiers, runs self-hosted so your provider keys stay local, and ships a live catalog that tracks which models are free this week — a landscape that changes constantly.
What do I get with Premium?
New free models, quota changes, and fixes reach your router the moment we ship them, instead of on the free 30-day delay. It's $19/year or $49 once.