Largest Context Window AI Models

A large context window lets you skip retrieval plumbing and just hand the model the whole document, codebase, or chat history. The trade is cost and, often, recall: a model that advertises a million-token window does not necessarily reason well across all of it. Rank by size here, then verify the model actually uses the far end of its window for your task before committing.

Showing top 25 of 694 matching models · last synced 8 hours ago

Top pick right now: Llama 4 Scout from OpenRouter. Its 10,000,000-token context window is the largest available right now.

# Model Provider Input $/Mtok Output $/Mtok Context Capabilities
1
Llama 4 Scout
OpenRouter $0.1 $0.3 10,000,000 vision, tools, structured output, open
2
Pareto Code Router
OpenRouter N/A N/A 2,000,000 N/A
3
Grok 4 Fast
OpenRouter $0.2 $0.5 2,000,000 vision, tools, reasoning, structured output
4
Grok 4.1 Fast
OpenRouter $0.2 $0.5 2,000,000 vision, tools, reasoning, structured output
9
Auto Router
OpenRouter N/A N/A 2,000,000 vision, tools, reasoning, structured output
10
Grok 4.20
OpenRouter $1.25 $2.5 2,000,000 vision, tools, reasoning, structured output
11
Grok 4.20 Multi-Agent
OpenRouter $1.25 $2.5 2,000,000 vision, reasoning, structured output
12
Grok 4.20 Multi - Agent Beta
OpenRouter $2 $6 2,000,000 vision, reasoning
13
Grok 4.20 Beta
OpenRouter $2 $6 2,000,000 vision, tools, reasoning
14
GPT-5.4 Pro
OpenRouter $30 $180 1,050,000 vision, tools, reasoning, structured output
15
Gemini 3 Pro Preview
OpenRouter $2 $12 1,050,000 vision, tools, reasoning, structured output
16
GPT-5.5
OpenRouter $5 $30 1,050,000 vision, tools, reasoning, structured output
17
GPT-5.4
OpenRouter $2.5 $15 1,050,000 vision, tools, reasoning, structured output
18
GPT-5.5 Pro
OpenRouter $30 $180 1,050,000 vision, tools, reasoning, structured output
23
OpenAI GPT Latest
OpenRouter $5 $30 1,050,000 vision, tools, reasoning, structured output

Browse other categories

Data synced daily from models.dev. Always verify pricing with the provider before deploying to production.