🤖 Generate website with AI 👆 →

Largest AI Context Windows

Grok 4.20 and Grok 4.20 Multi-Agent share the largest context window in this database: 2M tokens. Their context score is 79.

Data sourced from OpenRouter API and manual data collection | Last updated: Nov 3, 2026, 8:01 AM UTC

Ranked by context score

RankModelContextContext score
1SpaceXAI: Grok 4.202M79
1SpaceXAI: Grok 4.20 Multi-Agent2M79
3Meta: Llama 4 Scout1.31M74
4OpenAI: GPT-5.41.05M71
4OpenAI: GPT-5.4 Pro1.05M71
4OpenAI: GPT-5.51.05M71
4OpenAI: GPT-5.5 Pro1.05M71
4OpenAI: GPT-5.6 Luna1.05M71
4OpenAI: GPT-5.6 Luna Pro1.05M71
4OpenAI: GPT-5.6 Sol1.05M71
4OpenAI: GPT-5.6 Sol Pro1.05M71
4OpenAI: GPT-5.6 Terra1.05M71
4OpenAI: GPT-5.6 Terra Pro1.05M71
4OpenAI: GPT-6 Astra1.05M71
4OpenAI: GPT-6 Astra Pro1.05M71
4OpenAI: GPT-6 Luna1.05M71
4OpenAI: GPT-6 Luna Pro1.05M71
4OpenAI: GPT-6 Sol1.05M71
4OpenAI: GPT-6 Sol Pro1.05M71
4OpenAI: GPT-6.1 Sol1.05M71
4OpenAI: GPT-6.1 Sol Pro1.05M71
4OpenAI: GPT-6.1 Terra1.05M71
4OpenAI: GPT-6.1 Terra Pro1.05M71
4Xiaomi: MiMo-V2.51.05M71
4Xiaomi: MiMo-V2.5-Pro1.05M71
4Xiaomi: MiMo-V2.6-Flash1.05M71
4Xiaomi: MiMo-V2.6-Pro1.05M71
28Meituan: LongCat 2.01.05M71
29DeepSeek: DeepSeek V4 Flash 07311.05M71
29DeepSeek: DeepSeek V4 Flash Vision Exp1.05M71
29DeepSeek: DeepSeek V4 Pro 08131.05M71
29DeepSeek: DeepSeek V4.1 Flash 10171.05M71
29Fireworks: Ember-11.05M71
29Google: Gemini 2.5 Flash1.05M71
29Google: Gemini 2.5 Flash Lite1.05M71
29Google: Gemini 2.5 Pro1.05M71
29Google: Gemini 2.5 Pro Preview 06-051.05M71
29Google: Gemini 3 Flash Preview1.05M71
29Google: Gemini 3.1 Flash Lite1.05M71
29Google: Gemini 3.1 Flash Lite Preview1.05M71
29Google: Gemini 3.1 Pro Preview1.05M71
29Google: Gemini 3.1 Pro Preview Custom Tools1.05M71
29Google: Gemini 3.5 Flash1.05M71
29Google: Gemini 3.5 Flash Lite1.05M71
29Google: Gemini 3.6 Flash1.05M71
29Google: Gemini 3.7 Flash1.05M71
29Google: Gemini 3.8 Flash1.05M71
29Google: Gemini 4 Argon1.05M71
29Meta: Muse Spark 1.11.05M71
29Meta: Muse Spark 1.21.05M71
29Meta: Muse Spark 1.2 Contributor1.05M71
29Meta: Muse Spark 1.31.05M71
29Meta: Muse Spark 1.3 Contributor1.05M71
29Meta: Muse Spark Open1.05M71
29Meta: Muse Wonder1.05M71
29MiniMax: MiniMax M31.05M71
29MiniMax: MiniMax M3.1 Flash1.05M71
29Mistral: Mistral Large 41.05M71
29MoonshotAI: Kimi K31.05M71
29MoonshotAI: Kimi K3.1 Pro1.05M71
29NVIDIA: Nemotron 3 Super1.05M71
29NVIDIA: Nemotron 3.5 Super VL1.05M71
29Pareto 26.10: Pareto 26.101.05M71
29Particle AI: Particle-11.05M71
29Qwen: Qwen3.8 2.4T A95B1.05M71
29Reflection: Beam 1.0 501B A23B1.05M71
29StepFun: Step 5 Preview1.05M71
29Tencent: Hy4 preview1.05M71
29Xiaomi: MiMo-V2.6-Pro-UltraSpeed1.05M71
29Z.ai: GLM 5.21.05M71
29Z.ai: GLM 5.31.05M71
29Z.ai: GLM 5.3 Flash1.05M71
29Z.ai: GLM 5.3 FlashX1.05M71
29Z.ai: GLM 5.3 FlashX Legal1.05M71
75OpenAI: GPT-4.11.05M71
75OpenAI: GPT-4.1 Mini1.05M71
75OpenAI: GPT-4.1 Nano1.05M71
78Writer: Palmyra X51.04M71
79Qwen: Qwen3.8 27B1.02M71
79TheDrummer: UnslopNemo 12B1.02M71
81MiniMax: MiniMax-011M70
82Anthropic: Claude Fable 51M70
82Anthropic: Claude Fable 5.11M70
82Anthropic: Claude Haiku 5.51M70
82Anthropic: Claude Opus 4.61M70
82Anthropic: Claude Opus 4.71M70
82Anthropic: Claude Opus 4.81M70
82Anthropic: Claude Opus 51M70
82Anthropic: Claude Opus 5.51M70
82Anthropic: Claude Sonnet 4.51M70
82Anthropic: Claude Sonnet 4.61M70
82Anthropic: Claude Sonnet 51M70
82Anthropic: Claude Sonnet 5.51M70
82Inception: Mercury 31M70
82inclusionAI: Ling 3.1 Flash1M70
82MiniMax: MiniMax M11M70
82Qwen: Qwen-Plus1M70
82Qwen: Qwen3 Coder Flash1M70
82Qwen: Qwen3.5 Plus 2026-04-201M70
82Qwen: Qwen3.5-Flash1M70
82Qwen: Qwen3.7 Flash1M70
82Qwen: Qwen3.7 Max1M70
82Qwen: Qwen3.7 Plus1M70
82Qwen: Qwen3.8 Flash1M70
82Qwen: Qwen3.8 Max (0902)1M70
82Qwen: Qwen3.8 Max Prime1M70
82Qwen: Qwen3.8 Omni Flash1M70
82Sakana: Fugu Max1M70
82Sakana: Fugu Ultra1M70
82Sakana: Fugu Ultra v21M70
82SpaceXAI: Grok 4.31M70
82Z.ai: GLM 5.3 Prime1M70
113Meta: Llama 4 Maverick524K62
113Thinking Machines: Inkling524K62
113Thinking Machines: Inkling Small524K62
113Upstage: Solar Mini 4524K62
113Upstage: Solar Pro 4524K62
118SpaceXAI: Grok 4.5500K62
118SpaceXAI: Grok 4.6500K62
118SpaceXAI: Grok 4.7500K62
121OpenAI: GPT Chat Latest400K59
121OpenAI: GPT-5400K59
121OpenAI: GPT-5 Mini400K59
121OpenAI: GPT-5 Nano400K59
121OpenAI: GPT-5 Pro400K59
121OpenAI: GPT-5.1400K59
121OpenAI: GPT-5.1-Codex400K59
121OpenAI: GPT-5.1-Codex-Max400K59
121OpenAI: GPT-5.1-Codex-Mini400K59
121OpenAI: GPT-5.2400K59
121OpenAI: GPT-5.2 Pro400K59
121OpenAI: GPT-5.2-Codex400K59
121OpenAI: GPT-5.3-Codex400K59
121OpenAI: GPT-5.4 Mini400K59
121OpenAI: GPT-5.4 Nano400K59
136Amazon: Nova Lite 1.0300K55
136Amazon: Nova Pro 1.0300K55
138AionLabs: Aion 3.5262K53
138AionLabs: Aion 3.5 Mini262K53
138ByteDance Seed: Seed 1.6262K53
138ByteDance Seed: Seed 1.6 Flash262K53
138ByteDance Seed: Seed 2.1 Turbo262K53
138ByteDance Seed: Seed-2.0-Code262K53
138ByteDance Seed: Seed-2.0-Lite262K53
138ByteDance Seed: Seed-2.0-Mini262K53
138Elorian: Glimpse262K53
138Google: Gemma 4 26B A4B262K53
138Google: Gemma 4 31B262K53
138inclusionAI: Ling 3.0 Flash262K53
138Mistral: Devstral 2 2512262K53
138Mistral: Ministral 3 14B 2512262K53
138Mistral: Ministral 3 8B 2512262K53
138Mistral: Mistral Large 3 2512262K53
138Mistral: Mistral Medium 3.5262K53
138Mistral: Mistral Small 4262K53
138MoonshotAI: Kimi K2 0905262K53
138MoonshotAI: Kimi K2 Thinking262K53
138MoonshotAI: Kimi K2.6262K53
138MoonshotAI: Kimi K2.7 Code262K53
138Nex AGI: Nex-N2.5-Mini262K53
138Nex AGI: Nex-N2.5-Pro262K53
138NVIDIA: Nemotron 3 Nano 30B A3B262K53
138NVIDIA: Nemotron 3 Ultra262K53
138NVIDIA: Nemotron 3.5 Lightning262K53
138Pareto 26.9: Pareto 26.9262K53
138PrismML: Ternary Bonsai 2 27B262K53
138PrismML: Ternary Bonsai 2 9B262K53
138Qwen: Qwen3 235B A22B Instruct 2507262K53
138Qwen: Qwen3 30B A3B Instruct 2507262K53
138Qwen: Qwen3 Coder 30B A3B Instruct262K53
138Qwen: Qwen3 Coder 480B A35B262K53
138Qwen: Qwen3 Coder Next262K53
138Qwen: Qwen3 Next 80B A3B Instruct262K53
138Qwen: Qwen3 VL 235B A22B Instruct262K53
138Qwen: Qwen3 VL 30B A3B Instruct262K53
138Qwen: Qwen3 VL 30B A3B Thinking262K53
138Qwen: Qwen3 VL 8B Instruct262K53
138Qwen: Qwen3.5 397B A17B262K53
138Qwen: Qwen3.5-122B-A10B262K53
138Qwen: Qwen3.5-27B262K53
138Qwen: Qwen3.5-35B-A3B262K53
138Qwen: Qwen3.5-9B262K53
138Qwen: Qwen3.6 27B262K53
138Qwen: Qwen3.6 35B A3B262K53
138Sakana: Sakana Namazu262K53
138StepFun: Step 3.5 Flash262K53
138StepFun: Step 3.7 Flash262K53
138Tencent: Hy3262K53
138Tencent: Hy3 preview262K53
190Inception: Mercury 2.5260K53
191Cohere: Command A256K53
191Mistral: Codestral 2508256K53
191Mistral: Mistral Small 3.2 24B256K53
191SpaceXAI: Grok Build 0.1256K53
195MiniMax: MiniMax M2205K50
195MiniMax: MiniMax M2.5205K50
195MiniMax: MiniMax M2.7205K50
195Z.ai: GLM 4.6205K50
195Z.ai: GLM 4.7205K50
195Z.ai: GLM 5205K50
195Z.ai: GLM 5.1205K50
202Z.ai: GLM 5 Turbo203K50
202Z.ai: GLM 5V Turbo203K50
204Anthropic: Claude Haiku 4.5200K50
204Anthropic: Claude Opus 4.1200K50
204Anthropic: Claude Opus 4.5200K50
204OpenAI: o3200K50
204OpenAI: o3 Pro200K50
204Z.ai: GLM 4.7 Flash200K50
210DeepSeek: DeepSeek V3164K47
210DeepSeek: DeepSeek V3 0324164K47
210DeepSeek: DeepSeek V3.1164K47
210DeepSeek: DeepSeek V3.1 Terminus164K47
210DeepSeek: DeepSeek V3.2164K47
210DeepSeek: DeepSeek V3.2 Exp164K47
210DeepSeek: R1 0528164K47
210Meta: Llama Guard 4 12B164K47
218AionLabs: Aion-2.0131K44
218AionLabs: Aion-3.0131K44
218AionLabs: Aion-3.0-Mini131K44
218Google: Gemma 3 12B131K44
218Google: Gemma 3 27B131K44
218Google: Gemma 3 4B131K44
218IBM: Granite 4.2 8B131K44
218inclusionAI: Ling 3.0 Flash VL131K44
218Meta: Llama 3.1 70B Instruct131K44
218Meta: Llama 3.1 8B Instruct131K44
218Meta: Llama 3.2 3B Instruct131K44
218Meta: Llama 3.3 70B Instruct131K44
218Meta: Muse Glimmer 30B131K44
218Mistral Large 2407: Mistral Large 2407131K44
218Mistral: Ministral 3 3B 2512131K44
218Mistral: Mistral Medium 3131K44
218Mistral: Mistral Medium 3.1131K44
218Mistral: Mistral Nemo131K44
218MoonshotAI: Kimi K2 0711131K44
218Nous: Hermes 4 405B131K44
218OpenAI: gpt-oss-120b131K44
218OpenAI: gpt-oss-20b131K44
218OpenAI: gpt-oss-safeguard-20b131K44
218Qwen: Qwen3 32B131K44
218Sao10K: Llama 3.1 Euryale 70B v2.2131K44
218Tencent: Hunyuan A13B Instruct131K44
218TheDrummer: Cydonia 24B V4.1131K44
218Z.ai: GLM 4.5131K44
218Z.ai: GLM 4.5 Air131K44
247Amazon: Nova Micro 1.0128K44
247ByteDance: UI-TARS 7B128K44
247Cohere: Command R (08-2024)128K44
247Cohere: Command R+ (08-2024)128K44
247Cohere: Command R7B (12-2024)128K44
247Inception: Mercury 2128K44
247LiquidAI: LFM2.5-2.6B128K44
247LiquidAI: LFM2.5-24B-A2B128K44
247Mistral Large: Mistral Large128K44
247Mistral: Mistral Small 3.1 24B128K44
247OpenAI: GPT-4o128K44
247OpenAI: GPT-4o-mini128K44
247OpenAI: GPT-5.2 Chat128K44
247Qwen: Qwen2.5 VL 72B Instruct128K44
247Qwen: Qwen3 235B A22B Thinking 2507128K44
247Venice: Uncensored128K44
263MiniMax: MiniMax M2-her65.5K36
263Mistral: Mixtral 8x22B Instruct65.5K36
263Reka Flash 3: Reka Flash 365.5K36
263Z.ai: GLM 4.5V65.5K36
267Meta: Llama 3.2 1B Instruct60K34
268Qwen: Qwen3 14B41K30
268Qwen: Qwen3 30B A3B41K30
270Perceptron: Perceptron Mk1.536.9K28
271AionLabs: Aion-RP 1.0 (8B)32.8K27
271Magnum v4 72B: Magnum v4 72B32.8K27
271Mistral: Mistral Small 332.8K27
271Mistral: Saba32.8K27
271Mistral: Voxtral Small 24B 250732.8K27
271Perceptron: Perceptron Mk132.8K27
271Qwen: Qwen2.5 7B Instruct32.8K27
271Qwen2.5 72B Instruct: Qwen2.5 72B Instruct32.8K27
271Qwen2.5 Coder 32B Instruct: Qwen2.5 Coder 32B Instruct32.8K27
271TheDrummer: Skyfall 36B V232.8K27
281Moondream: Moondream 3 Preview32K26
281Moondream: Moondream 3.132K26
283OpenAI: GPT-3.5 Turbo 16k16.4K18
284Microsoft: Phi 416.4K18
285Google: Gemma 2 27B8.2K9
285MythoMax 13B: MythoMax 13B8.2K9
285Sao10K: Llama 3 8B Lunaris8.2K9
288OpenAI: GPT-48.2K9
289Mancer: Weaver (alpha)8K9
290ReMM SLERP 13B: ReMM SLERP 13B6.1K5
291OpenAI: GPT-3.5 Turbo (older v0613)4.1K0

🤖 Generate website with AI 👆

Join the beta waitlist and get 50% off your first 3 months.

Join waitlist

Frequently asked questions

Which AI model has the largest context window?

Grok 4.20 and Grok 4.20 Multi-Agent share the largest context window in this database: 2M tokens. Their context score is 79.

How is the context score calculated?

The context score places the window on a 100-point log scale between 4,096 and 10,000,000 tokens. A larger window scores higher.