LLM API Price Comparison

Compare API prices for leading models like Claude, GPT, Gemini, and Grok to find a provider that fits your budget.

Listed prices by model

Rankings updated: All providers. 29 providers match.

Anthropic

Anthropic Claude Opus 5.5
Anthropic Claude Opus 5
Anthropic Claude Fable 5.2
Anthropic Claude Fable 5.1

OpenAI

OpenAI GPT 6.1 Sol
OpenAI GPT 6.1 Terra
OpenAI GPT 6 Astra
OpenAI GPT 5.6 Sol

Google

Google Gemini 4 Hydrogen
Google Gemini 4 Carbon
Google Gemini 4 Argon
Google Gemini 3.8 Flash

xAI

xAI Grok 4.7
xAI Grok 4.6
xAI Grok 4.5

DeepSeek

DeepSeek DeepSeek V4 Pro
DeepSeek DeepSeek V4.1 Flash
DeepSeek DeepSeek V4 Flash

Moonshot

Moonshot Kimi k3.1 Pro
Moonshot Kimi K3

How to use the price ranking

  1. Pick a model

    Each board compares listed prices for one model. Switch models from the name on the card.

  2. Read the order and the percentage

    The price ranking sorts providers by how many tokens $100 buys at a 1:1 mix. The percentage is versus the official listed price.

  3. Open the provider page

    The ranking is not the amount you pay. Check the input and output prices on the provider page before you pay.

Submit a public price for review

A listing requires a public pricing page where the input and output rates can be checked again.

Listed price questions

How the ranking is calculated, where prices come from, and what the homepage leaves out.

How are the price rankings sorted?

Each board is one model. Providers are sorted by how many tokens $100 buys at the listed input and output prices, using a 1:1 mix. The percentage compares that price with the official listed price. The bar compares each provider with first place on that board.

Why does the same provider rank differently by model?

Input and output prices are separate, so a provider can lead on Claude and fall behind on GPT or Gemini. A provider appears only when it publishes a complete listed price for that exact model. Missing quotes stay blank.

Is the savings percentage the price you pay?

No. The percentage compares listed input and output prices with the official price. It is not your invoice. It also does not measure uptime, routing, privacy, refunds, or whether the model is genuine. Open the provider page before you pay.

Do cache, plans, and top-up bonuses affect the rank?

No. The rank uses listed input and output prices only. Cache rates, plans, top-up bonuses, and payment fees stay out.

Where do the listed prices come from, and how often are they checked?

Provider prices come from public pricing pages and are stored with the source link and observation date. Official prices come from the model vendor. Prices change, so check the provider page before you pay.

How is the homepage different from a single model page?

The homepage scans several models at $100 and a balanced 1:1 mix. A model page compares one model, lets you change the budget and mix, and lists the full public quotes for that model.

Submit a site