AI Models Compared: Context Window and Price List

New AI models arrive every week and the names are hard to track. This page lists the major models from OpenAI, Anthropic, Google, Meta, DeepSeek and others in one table, with how much text each can read at once, what it costs per million tokens and what kinds of input it accepts. The numbers load live from a public model catalog. This is a fact sheet, not a quality ranking: a bigger window or a lower price does not make a model better for your task. To try the tools built on these models, see the AI tools directory, or look up a term in the AI glossary.

The model list loads below from OpenRouter's public catalog.

How to read this table

  • Context window is how much text the model can take in at one time, counted in tokens. A token is a few letters, so 1,000 tokens is roughly 750 English words.
  • Price per million tokens is split into input (what you send) and output (what the model writes). Output usually costs more.
  • Inputs show what the model can read: text only, or also images, audio, video or files.
  • The date is when the model was added to the catalog, which is close to its release.

Picking a model for a job

Long documents

Look for a large context window, then test whether the model keeps track of details in the middle of a long file.

High volume or low budget

Compare the output price first, since long answers add up. Smaller models in each family are often far cheaper and good enough for simple tasks.

Images, audio or video

Check the inputs column. Not every strong text model can read pictures or sound.

Hard reasoning or coding

Price and window say little here. Try two or three models on your own examples and compare the answers.

Questions people ask

Where do these numbers come from?

From OpenRouter's public model catalog. Each row links to the model's page there. Makers may list different prices on their own sites, so check before you budget.

How often does the list update?

About once an hour.

Is this a leaderboard?

No. Rankings depend on the task and change quickly. This page only compares facts that are easy to measure.

Why is a model missing?

We show major text models from well-known labs. Image-only, voice-only and retired models are left out.

What is a token?

A small piece of text, often part of a word. Models read and write in tokens, and prices are set per token. Our AI glossary has more terms.