AI and LLMs · Comparison

OpenAI vs Gemini vs Claude vs OpenRouter

Three model makers and one gateway in front of them all. Prices are rough ranges per million input tokens; output tokens cost about four to six times more.

4 options · 8 questions side by side · updated

CompareOpenAIGeminiClaudeOpenRouter
What it isOpenAI's own modelsGoogle's own modelsAnthropic's own modelsOne API for hundreds of models
Model tiersSmall, mid-size and flagship reasoningFlash-Lite, Flash and ProHaiku, Sonnet, Opus and aboveWhatever each provider offers
Free tierNoYes, with rate limitsNoFree models with daily limits
Input price per millionAbout $0.10 to $10About $0.10 to $2 (more for long prompts)About $1 to $10Provider price, plus a fee on credits
Long contextYes, priced higher past about 272K tokensUp to about 1M tokensUp to about 1M tokens on recent modelsDepends on the model
Stand-out extrasImage, speech and realtime voice modelsSearch grounding, video and audio inputComputer use, citations, prompt cachingRouting, fallbacks, one bill
Your dataNot used for training by defaultFree-tier prompts may improve Google productsNot used for training by defaultSkips hosts that train on prompts by default
Watch out forModel retirements force upgradesDifferent terms for free and paid usePricier small models than rivalsFee on credits and one more middleman

How to choose between OpenAI, Gemini, Claude and OpenRouter

  • Pick OpenAI for a wide range of models and mature tooling across text, images and speech.
  • Pick Gemini to prototype for free, or for long documents, audio and video at low per-token prices.
  • Pick Claude for coding assistants, long documents and agents that lean heavily on tools.
  • Pick OpenRouter to compare models, or to mix providers behind one key and one bill.

The options

More comparisons

Crafted in the dark. Shipped to the world.

Tell us what you are building. You get a private project space with a proposal and a line-by-line quote within a day.