Skip to content

DirectoryTools

LLMFly AI logo

LLMFly AI

GPT, Claude, Gemini and Grok behind one OpenAI-compatible API.

Visit LLMFly AI

Coding & WebFreemiumWe checked the prices on llmfly.ai on 27 Sep 2026.

LLMFly AI homepage screenshot

What is LLMFly AI?

GPT, Claude, Gemini and Grok behind one OpenAI-compatible API.

Who is LLMFly AI best for?

LLMFly AI is best for developers who want GPT, Claude, Gemini and Grok models through one key.

  • Apps that call several providers

    GPT, Claude, Gemini and Grok models sit behind one OpenAI-compatible endpoint, so you change the base URL and model ID instead of adding a new SDK for each.

  • Coding agents on a budget

    Setup guides cover Claude Code, Codex, Cursor and Cherry Studio, and GPT models bill at 0.3x OpenAI's list price.

  • Separate keys per project

    Each app or environment can get its own key, so a leaked key can be revoked without taking the others offline, and the usage log shows what every request cost.

How much does LLMFly AI cost?

LLMFly AI is pay as you go: you prepay credits and each model has its own rate, from $0.04 per million input tokens for Gemini 2.0 Flash to $5 for Claude Fable 5.1, with some free credits when you sign up.

Gemini modelsFrom $0.04Per million input tokens
  • 8 models at 0.4x Google's list price
  • Gemini 3.8 Flash: $0.30 in, $1.50 out per million tokens
  • Gemini 3 Pro Preview: $0.80 in, $4.80 out up to 200K tokens
ChatGPT (Codex) modelsFrom $0.06Per million input tokens
  • 5 models at 0.3x OpenAI's list price
  • GPT-6 Astra: $3 in, $15 out per million tokens
  • GPT-5.6 Terra: $0.60 in, $3.60 out up to 272K tokens
Claude Code modelsFrom $0.50Per million input tokens
  • 13 models at 0.5x Anthropic's list price
  • Claude Fable 5.1: $5 in, $25 out per million tokens
  • Claude Sonnet 5: $1 in, $5 out; Claude Haiku 4.5: $0.50 in, $2.50 out
Grok modelsFrom $0.80Per million input tokens
  • 2 models at 0.4x xAI's list price
  • Grok 4.6: $0.80 in, $2.40 out under 200K tokens
  • $1.60 in, $4.80 out for longer requests

All rates are in US dollars per million tokens and already include each group's multiplier; cache reads, cache writes and tool calls are billed separately. Credits are prepaid, have no cash value and are not refundable except for billing errors. New accounts get free credits, though the amount is not stated, and referrals earn 10%. The live rate for each model is in the Model Plaza after you sign in. Full pricing on llmfly.ai. We checked these prices on .

How do readers rate LLMFly AI?

No one has rated LLMFly AI yet. Be the first.

No ratings yetThe score and its stars show once 3 readers have rated LLMFly AI.
Your rating

Sign in to rate. One rating per account.

Your rating of LLMFly AI

What are the best LLMFly AI alternatives?

One tool on Recaply does LLMFly AI's job. It is open source, so you can run it on your own server.

Questions about LLMFly AI

What is LLMFly AI?

GPT, Claude, Gemini and Grok behind one OpenAI-compatible API.

Who is LLMFly AI for?

LLMFly AI is best for developers who want GPT, Claude, Gemini and Grok models through one key.

Is LLMFly AI free?

No. LLMFly AI is a pay-as-you-go API: you prepay credits and every request is charged at that model's rate. New accounts get some free credits to test with, but LLMFly does not say how many, and prepaid credits are not refundable.

How much does LLMFly AI cost?

LLMFly AI's paid plans are Gemini models at From $0.04 per million input tokens, ChatGPT (Codex) models at From $0.06 per million input tokens, Claude Code models at From $0.50 per million input tokens and Grok models at From $0.80 per million input tokens.

Is there an open source alternative to LLMFly AI?

Yes. APIPark is an open source tool that does the same job, so you can run your own.

What are the best LLMFly AI alternatives?

The closest alternatives on Recaply are APIPark.

Where can I try LLMFly AI?

On its official website, llmfly.ai.

Is LLMFly AI an official OpenAI or Anthropic API?

No. LLMFly AI is a relay run by Flamewright Technologies Ltd, a company registered in England and Wales. It forwards requests to GPT, Claude, Gemini and Grok models through its own endpoint and bills them at its own rates, which it lists next to each provider's official price.