Which LLM has the longest context?

See the live table. Catalog max_context changes as new rows appear.

Is 1M context worth it?

Only if you fill it with useful tokens and the model still attends. Price it first.

Does longer context mean smarter?

No. It means a higher cap. Quality index is a separate column.

How do I rank by context here?

Set the context weight to 100% on the LLM leaderboard or use this page’s table.

Longest context window LLM — who actually fits the doc

The longest context window LLM is a spec race that you can lose on the invoice. A million tokens of context is a million tokens of input billing (plus worse latency). Buy the window you will fill with high-signal tokens.

This page’s live table ranks the catalog by max context. Click through to see $ / 1M. If the long-context row is expensive, a RAG pipeline on a 128k model is usually the adult architecture.

Longest context windows right now

Live OpenRouter pricing and Arena quality, cached about an hour. Not a static blog table.

Model	Provider	Blend $ / 1M	Quality	Context
Grok 4.20 Multi-Agent	xAI	1.88	—	2,000,000
Grok 4.20	xAI	1.88	—	2,000,000
Auto Router (Beta)	OpenRouter	—	—	2,000,000
Pareto Code Router	OpenRouter	—	—	2,000,000
Auto Router	OpenRouter	—	—	2,000,000

Marketing windows vs usable windows

Some models advertise huge contexts but degrade (“lost in the middle”). Treat the catalog number as a hard cap, not as quality at that cap. Needle tests on your documents are required if you plan to stuff books.

Cost of a full window

Filling 200k tokens at $0.50/M input is $0.10 per request before the answer. At 100k requests, that is $10k. The longest window is not free capacity; it is a pricing dimension.

Output limits

Max output is separate. You can have a million-token input and a 4k output cap. Check both fields on the model page if you want long generations.

When long context is the right tool

One-off legal or research docs, messy exports, or when retrieval quality is poor and you can afford brute force. Not for a chatbot that should have looked up one article.

Gemini and others

Google has competed hard on long context; others have followed. Do not assume the leader this quarter without sorting the live table. Pair with GPT vs Gemini if that is your shortlist.

FAQ

Which LLM has the longest context?: See the live table. Catalog max_context changes as new rows appear.
Is 1M context worth it?: Only if you fill it with useful tokens and the model still attends. Price it first.
Does longer context mean smarter?: No. It means a higher cap. Quality index is a separate column.
How do I rank by context here?: Set the context weight to 100% on the LLM leaderboard or use this page’s table.

longest context window LLM1 million token contextlargest context AI modellong context LLM comparison

Rank by max context