FreeLLMAPI

Author codeRa1nTotal Visits 2Visits in 7 Days 0

// preview

// description

An OpenAI-compatible gateway that routes to many free LLM providers with one API key

FreeLLMAPI is a developer dashboard that unifies many free and free-tier LLM providers behind a single OpenAI-compatible endpoint. After signing in you can add provider credentials from a catalog of 57 platforms, choose a routing strategy such as Manual, Balanced, Smartest, Fastest or Most reliable, manage named fallback chains, test models in a chat playground with sampling controls, and generate configuration for coding agents including Claude Code, Codex CLI, Cline and OpenCode. An analytics area reports request volume, latency, token usage, errors and estimated savings.

Highlights

One unified API key

A single endpoint for many providers

Six routing strategies

Manual through most reliable

57-provider catalog

Free and free-tier model sources

Chat playground

Test models and sampling live

Agent setup recipes

Claude Code, Codex, Cline and more

Usage analytics

Latency, tokens, errors and savings

How to Use

Sign in with the demo account
Open the site and you are taken to a sign-in screen. The page prints a public demo account, and the email and password fields are already filled in, so you can press Sign in to reach the dashboard immediately.
Add a provider key
Go to the Keys page and press Add key. In the dialog choose Provider key, open the Platform dropdown, filter or scroll the catalog, select a provider, paste its API key, optionally add a label, and confirm. The page then reports how many of the 57 providers are configured.
Choose how requests route
On the Models page pick a routing strategy. Manual lets you drag to set the order, while the other options route by live score across reliability, speed and intelligence. You can also open the Fallback chains panel to define named chains that are callable as auto entries.
Test a model in the Playground
Open the Playground, select a model in the right-hand Settings column, optionally write a system prompt, switch on temperature or top p if you want to override defaults, then type a message and send it to see the response in the conversation area.
Connect an app or coding agent
The Agents page lists supported tools. For each one you can copy an automatic setup command, or copy the base URL and unified API key to enter manually in that tool's own settings, then follow the linked documentation.
Review analytics
The Analytics page shows metric tiles for requests, success rate, input and output tokens, average and p95 latency, average time to first token and estimated savings. Use the time-range tabs and the calls table filters to narrow the view.