Free LLM API

How to evaluate free LLM API access

A free LLM API may mean a genuinely zero-priced model, temporary trial credits, a limited free tier, or a self-hosted model with separate infrastructure costs. Check the billing unit and usage limits before sending production traffic.

Maintained by LivRouter. Reviewed . Prices and availability can change; verify the live catalog before deployment.

Zero-priced models in the LivRouter catalog

The current catalog has no model with both zero-priced input and output usage. LivRouter's Start option is pay-as-you-go, not a free tier. The catalog is checked at render time and may change.

Source: LivRouter model catalog. Checked .

Questions to check before choosing a free API

Is access actually free?
Separate zero-priced usage from signup credits or a limited promotion. Confirm whether input, output, and cached tokens are all covered.
What are the limits?
Review request and token caps, throttling, context size, model availability, and whether the provider can change or end the offer.
Can it handle your workload?
Test representative prompts and verify latency, structured output or tool support, data handling terms, and reliability before relying on a free endpoint.
What does self-hosting cost?
Open weights may avoid per-token API charges, but compute, storage, operations, and maintenance are still costs.