Compare 287 AI models with context windows of 32K tokens or more. The largest models support 2M tokens -- enough to process entire codebases, books, or hundreds of documents in a single prompt. Data updated hourly.
The largest context windows available. These models can process entire codebases, full books, or hundreds of documents in a single request.
| # | Model | Provider | Context Window | Pages | Input / 1M | Output / 1M |
|---|---|---|---|---|---|---|
| 1 | Grok 4 Fast | xAI | 2M | ~3K | $0.20 | $0.50 |
| 2 | Grok 4.1 Fast | xAI | 2M | ~3K | $0.20 | $0.50 |
| 3 | Grok 4.20 Beta | xAI | 2M | ~3K | $2.00 | $6.00 |
| 4 | Grok 4.20 Multi-Agent Beta | xAI | 2M | ~3K | $2.00 | $6.00 |
| 5 | GPT-5.4 | OpenAI | 1.1M | ~1K | $2.50 | $15.00 |
| 6 | GPT-5.4 Pro | OpenAI | 1.1M | ~1K | $30.00 | $180.00 |
| 7 | Gemini 2.0 Flash | 1.0M | ~1K | $0.10 | $0.40 | |
| 8 | Gemini 2.0 Flash Lite | 1.0M | ~1K | $0.07 | $0.30 | |
| 9 | Gemini 2.5 Flash | 1.0M | ~1K | $0.30 | $2.50 | |
| 10 | Gemini 2.5 Flash Lite | 1.0M | ~1K | $0.10 | $0.40 | |
| 11 | Gemini 2.5 Flash Lite Preview 09-2025 | 1.0M | ~1K | $0.10 | $0.40 | |
| 12 | Gemini 2.5 Pro | 1.0M | ~1K | $1.25 | $10.00 | |
| 13 | Gemini 2.5 Pro Preview 05-06 | 1.0M | ~1K | $1.25 | $10.00 | |
| 14 | Gemini 2.5 Pro Preview 06-05 | 1.0M | ~1K | $1.25 | $10.00 | |
| 15 | Gemini 3 Flash Preview | 1.0M | ~1K | $0.50 | $3.00 | |
| 16 | Gemini 3 Pro Preview | 1.0M | ~1K | $2.00 | $12.00 | |
| 17 | Gemini 3.1 Flash Lite Preview | 1.0M | ~1K | $0.25 | $1.50 | |
| 18 | Gemini 3.1 Pro Preview | 1.0M | ~1K | $2.00 | $12.00 | |
| 19 | Gemini 3.1 Pro Preview Custom Tools | 1.0M | ~1K | $2.00 | $12.00 | |
| 20 | Llama 4 Maverick | Meta | 1.0M | ~1K | $0.15 | $0.60 |
| 21 | MiMo-V2-Pro | Xiaomi | 1.0M | ~1K | $1.00 | $3.00 |
| 22 | GPT-4.1 | OpenAI | 1.0M | ~1K | $2.00 | $8.00 |
| 23 | GPT-4.1 Mini | OpenAI | 1.0M | ~1K | $0.40 | $1.60 |
| 24 | GPT-4.1 Nano | OpenAI | 1.0M | ~1K | $0.10 | $0.40 |
| 25 | Palmyra X5 | Writer | 1.0M | ~1K | $0.60 | $6.00 |
| 26 | MiniMax-01 | MiniMax | 1.0M | ~1K | $0.20 | $1.10 |
| 27 | Claude Opus 4.6 | Anthropic | 1M | ~1K | $5.00 | $25.00 |
| 28 | Claude Sonnet 4.5 | Anthropic | 1M | ~1K | $3.00 | $15.00 |
| 29 | Claude Sonnet 4.6 | Anthropic | 1M | ~1K | $3.00 | $15.00 |
| 30 | MiniMax M1 | MiniMax | 1M | ~1K | $0.40 | $2.20 |
| 31 | Nova 2 Lite | Amazon | 1M | ~1K | $0.30 | $2.50 |
| 32 | Nova Premier 1.0 | Amazon | 1M | ~1K | $2.50 | $12.50 |
| 33 | Qwen Plus 0728 | Alibaba | 1M | ~1K | $0.26 | $0.78 |
| 34 | Qwen Plus 0728 (thinking) | Alibaba | 1M | ~1K | $0.26 | $0.78 |
| 35 | Qwen-Plus | Alibaba | 1M | ~1K | $0.26 | $0.78 |
| 36 | Qwen3 Coder Flash | Alibaba | 1M | ~1K | $0.20 | $0.97 |
| 37 | Qwen3 Coder Plus | Alibaba | 1M | ~1K | $0.65 | $3.25 |
| 38 | Qwen3.5 Plus 2026-02-15 | Alibaba | 1M | ~1K | $0.26 | $1.56 |
| 39 | Qwen3.5-Flash | Alibaba | 1M | ~1K | $0.07 | $0.26 |
Extended context models ideal for long documents, legal contracts, research papers, and multi-file code analysis.
| # | Model | Provider | Context Window | Pages | Input / 1M | Output / 1M |
|---|---|---|---|---|---|---|
| 1 | GPT-5 | OpenAI | 400K | ~533 | $1.25 | $10.00 |
| 2 | GPT-5 Codex | OpenAI | 400K | ~533 | $1.25 | $10.00 |
| 3 | GPT-5 Image | OpenAI | 400K | ~533 | $10.00 | $10.00 |
| 4 | GPT-5 Image Mini | OpenAI | 400K | ~533 | $2.50 | $2.00 |
| 5 | GPT-5 Mini | OpenAI | 400K | ~533 | $0.25 | $2.00 |
| 6 | GPT-5 Nano | OpenAI | 400K | ~533 | $0.05 | $0.40 |
| 7 | GPT-5 Pro | OpenAI | 400K | ~533 | $15.00 | $120.00 |
| 8 | GPT-5.1 | OpenAI | 400K | ~533 | $1.25 | $10.00 |
| 9 | GPT-5.1-Codex | OpenAI | 400K | ~533 | $1.25 | $10.00 |
| 10 | GPT-5.1-Codex-Max | OpenAI | 400K | ~533 | $1.25 | $10.00 |
| 11 | GPT-5.1-Codex-Mini | OpenAI | 400K | ~533 | $0.25 | $2.00 |
| 12 | GPT-5.2 | OpenAI | 400K | ~533 | $1.75 | $14.00 |
| 13 | GPT-5.2 Pro | OpenAI | 400K | ~533 | $21.00 | $168.00 |
| 14 | GPT-5.2-Codex | OpenAI | 400K | ~533 | $1.75 | $14.00 |
| 15 | GPT-5.3-Codex | OpenAI | 400K | ~533 | $1.75 | $14.00 |
| 16 | GPT-5.4 Mini | OpenAI | 400K | ~533 | $0.75 | $4.50 |
| 17 | GPT-5.4 Nano | OpenAI | 400K | ~533 | $0.20 | $1.25 |
| 18 | Llama 4 Scout | Meta | 328K | ~437 | $0.08 | $0.30 |
| 19 | Nova Lite 1.0 | Amazon | 300K | ~400 | $0.06 | $0.24 |
| 20 | Nova Pro 1.0 | Amazon | 300K | ~400 | $0.80 | $3.20 |
| 21 | Devstral 2 2512 | Mistral AI | 262K | ~350 | $0.40 | $2.00 |
| 22 | Kimi K2.5 | Moonshot AI | 262K | ~350 | $0.45 | $2.20 |
| 23 | MiMo-V2-Flash | Xiaomi | 262K | ~350 | $0.09 | $0.29 |
| 24 | MiMo-V2-Omni | Xiaomi | 262K | ~350 | $0.40 | $2.00 |
| 25 | Ministral 3 14B 2512 | Mistral AI | 262K | ~350 | $0.20 | $0.20 |
| 26 | Ministral 3 8B 2512 | Mistral AI | 262K | ~350 | $0.15 | $0.15 |
| 27 | Mistral Large 3 2512 | Mistral AI | 262K | ~350 | $0.50 | $1.50 |
| 28 | Mistral Small 4 | Mistral AI | 262K | ~350 | $0.15 | $0.60 |
| 29 | Nemotron 3 Nano 30B A3B | NVIDIA | 262K | ~350 | $0.05 | $0.20 |
| 30 | Nemotron 3 Super | NVIDIA | 262K | ~350 | $0.10 | $0.50 |
| 31 | Nemotron 3 Super (free) | NVIDIA | 262K | ~350 | Free | Free |
| 32 | Qwen3 235B A22B Instruct 2507 | Alibaba | 262K | ~350 | $0.07 | $0.10 |
| 33 | Qwen3 30B A3B Instruct 2507 | Alibaba | 262K | ~350 | $0.09 | $0.30 |
| 34 | Qwen3 Coder 480B A35B | Alibaba | 262K | ~350 | $0.22 | $1.00 |
| 35 | Qwen3 Coder Next | Alibaba | 262K | ~350 | $0.12 | $0.75 |
| 36 | Qwen3 Max | Alibaba | 262K | ~350 | $0.78 | $3.90 |
| 37 | Qwen3 Max Thinking | Alibaba | 262K | ~350 | $0.78 | $3.90 |
| 38 | Qwen3 Next 80B A3B Instruct | Alibaba | 262K | ~350 | $0.09 | $1.10 |
| 39 | Qwen3 Next 80B A3B Instruct (free) | Alibaba | 262K | ~350 | Free | Free |
| 40 | Qwen3 VL 235B A22B Instruct | Alibaba | 262K | ~350 | $0.20 | $0.88 |
| 41 | Qwen3.5 397B A17B | Alibaba | 262K | ~350 | $0.39 | $2.34 |
| 42 | Qwen3.5-122B-A10B | Alibaba | 262K | ~350 | $0.26 | $2.08 |
| 43 | Qwen3.5-27B | Alibaba | 262K | ~350 | $0.20 | $1.56 |
| 44 | Qwen3.5-35B-A3B | Alibaba | 262K | ~350 | $0.16 | $1.30 |
| 45 | Seed 1.6 | ByteDance | 262K | ~350 | $0.25 | $2.00 |
| 46 | Seed 1.6 Flash | ByteDance | 262K | ~350 | $0.07 | $0.30 |
| 47 | Seed-2.0-Lite | ByteDance | 262K | ~350 | $0.25 | $2.00 |
| 48 | Seed-2.0-Mini | ByteDance | 262K | ~350 | $0.10 | $0.40 |
| 49 | Qwen3 Coder 480B A35B (free) | Alibaba | 262K | ~349 | Free | Free |
| 50 | Codestral 2508 | Mistral AI | 256K | ~341 | $0.30 | $0.90 |
| 51 | Command A | Cohere | 256K | ~341 | $2.50 | $10.00 |
| 52 | Grok 4 | xAI | 256K | ~341 | $3.00 | $15.00 |
| 53 | Grok Code Fast 1 | xAI | 256K | ~341 | $0.20 | $1.50 |
| 54 | Jamba Large 1.7 | AI21 Labs | 256K | ~341 | $2.00 | $8.00 |
| 55 | KAT-Coder-Pro V1 | Kuaishou | 256K | ~341 | $0.21 | $0.83 |
| 56 | Nemotron 3 Nano 30B A3B (free) | NVIDIA | 256K | ~341 | Free | Free |
| 57 | Qwen3.5-9B | Alibaba | 256K | ~341 | $0.05 | $0.15 |
| 58 | Step 3.5 Flash | StepFun | 256K | ~341 | $0.10 | $0.30 |
| 59 | Step 3.5 Flash (free) | StepFun | 256K | ~341 | Free | Free |
| 60 | MiniMax M2.7 | MiniMax | 205K | ~273 | $0.30 | $1.20 |
| 61 | Claude 3 Haiku | Anthropic | 200K | ~267 | $0.25 | $1.25 |
| 62 | Claude 3.5 Haiku | Anthropic | 200K | ~267 | $0.80 | $4.00 |
| 63 | Claude 3.5 Sonnet | Anthropic | 200K | ~267 | $6.00 | $30.00 |
| 64 | Claude 3.7 Sonnet | Anthropic | 200K | ~267 | $3.00 | $15.00 |
| 65 | Claude 3.7 Sonnet (thinking) | Anthropic | 200K | ~267 | $3.00 | $15.00 |
| 66 | Claude Haiku 4.5 | Anthropic | 200K | ~267 | $1.00 | $5.00 |
| 67 | Claude Opus 4 | Anthropic | 200K | ~267 | $15.00 | $75.00 |
| 68 | Claude Opus 4.1 | Anthropic | 200K | ~267 | $15.00 | $75.00 |
| 69 | Claude Opus 4.5 | Anthropic | 200K | ~267 | $5.00 | $25.00 |
| 70 | Claude Sonnet 4 | Anthropic | 200K | ~267 | $3.00 | $15.00 |
| 71 | Composer 2 | Cursor | 200K | ~267 | $0.50 | $2.50 |
| 72 | Composer 2 Fast | Cursor | 200K | ~267 | $1.50 | $7.50 |
| 73 | o1 | OpenAI | 200K | ~267 | $15.00 | $60.00 |
| 74 | o1-pro | OpenAI | 200K | ~267 | $150.00 | $600.00 |
| 75 | o3 | OpenAI | 200K | ~267 | $2.00 | $8.00 |
| 76 | o3 Deep Research | OpenAI | 200K | ~267 | $10.00 | $40.00 |
| 77 | o3 Mini | OpenAI | 200K | ~267 | $1.10 | $4.40 |
| 78 | o3 Mini High | OpenAI | 200K | ~267 | $1.10 | $4.40 |
| 79 | o3 Pro | OpenAI | 200K | ~267 | $20.00 | $80.00 |
| 80 | o4 Mini | OpenAI | 200K | ~267 | $1.10 | $4.40 |
| 81 | o4 Mini Deep Research | OpenAI | 200K | ~267 | $2.00 | $8.00 |
| 82 | o4 Mini High | OpenAI | 200K | ~267 | $1.10 | $4.40 |
| 83 | Sonar Pro | Perplexity | 200K | ~267 | $3.00 | $15.00 |
| 84 | Sonar Pro Search | Perplexity | 200K | ~267 | $3.00 | $15.00 |
The current standard for frontier models. Sufficient for most production use cases, including long conversations and medium-length documents.
| # | Model | Provider | Context Window | Pages | Input / 1M | Output / 1M |
|---|---|---|---|---|---|---|
| 1 | MiniMax M2 | MiniMax | 197K | ~262 | $0.26 | $1.00 |
| 2 | MiniMax M2.1 | MiniMax | 197K | ~262 | $0.27 | $0.95 |
| 3 | MiniMax M2.5 | MiniMax | 197K | ~262 | $0.20 | $1.17 |
| 4 | MiniMax M2.5 (free) | MiniMax | 197K | ~262 | Free | Free |
| 5 | DeepSeek V3 | DeepSeek | 164K | ~218 | $0.32 | $0.89 |
| 6 | DeepSeek V3 0324 | DeepSeek | 164K | ~218 | $0.20 | $0.77 |
| 7 | DeepSeek V3.1 Terminus | DeepSeek | 164K | ~218 | $0.21 | $0.79 |
| 8 | DeepSeek V3.2 | DeepSeek | 164K | ~218 | $0.26 | $0.38 |
| 9 | DeepSeek V3.2 Exp | DeepSeek | 164K | ~218 | $0.27 | $0.41 |
| 10 | DeepSeek V3.2 Speciale | DeepSeek | 164K | ~218 | $0.40 | $1.20 |
| 11 | Llama Guard 4 12B | Meta | 164K | ~218 | $0.18 | $0.18 |
| 12 | R1 0528 | DeepSeek | 164K | ~218 | $0.45 | $2.15 |
| 13 | Qwen3 Coder 30B A3B Instruct | Alibaba | 160K | ~213 | $0.07 | $0.27 |
| 14 | Aion-1.0 | aion-labs | 131K | ~175 | $4.00 | $8.00 |
| 15 | Aion-1.0-Mini | aion-labs | 131K | ~175 | $0.70 | $1.40 |
| 16 | Aion-2.0 | aion-labs | 131K | ~175 | $0.80 | $1.60 |
| 17 | Devstral Medium | Mistral AI | 131K | ~175 | $0.40 | $2.00 |
| 18 | Devstral Small 1.1 | Mistral AI | 131K | ~175 | $0.10 | $0.30 |
| 19 | ERNIE 4.5 21B A3B Thinking | Baidu | 131K | ~175 | $0.07 | $0.28 |
| 20 | Gemma 3 12B | 131K | ~175 | $0.04 | $0.13 | |
| 21 | Gemma 3 27B | 131K | ~175 | $0.08 | $0.16 | |
| 22 | Gemma 3 27B (free) | 131K | ~175 | Free | Free | |
| 23 | Gemma 3 4B | 131K | ~175 | $0.04 | $0.08 | |
| 24 | gpt-oss-120b | OpenAI | 131K | ~175 | $0.04 | $0.19 |
| 25 | gpt-oss-120b (free) | OpenAI | 131K | ~175 | Free | Free |
| 26 | gpt-oss-20b | OpenAI | 131K | ~175 | $0.03 | $0.11 |
| 27 | gpt-oss-20b (free) | OpenAI | 131K | ~175 | Free | Free |
| 28 | gpt-oss-safeguard-20b | OpenAI | 131K | ~175 | $0.07 | $0.30 |
| 29 | Grok 3 | xAI | 131K | ~175 | $3.00 | $15.00 |
| 30 | Grok 3 Beta | xAI | 131K | ~175 | $3.00 | $15.00 |
| 31 | Grok 3 Mini | xAI | 131K | ~175 | $0.30 | $0.50 |
| 32 | Grok 3 Mini Beta | xAI | 131K | ~175 | $0.30 | $0.50 |
| 33 | Hunyuan A13B Instruct | Tencent | 131K | ~175 | $0.14 | $0.57 |
| 34 | Kimi K2 0905 | Moonshot AI | 131K | ~175 | $0.40 | $2.00 |
| 35 | Kimi K2 Thinking | Moonshot AI | 131K | ~175 | $0.47 | $2.00 |
| 36 | Llama 3.1 70B Instruct | Meta | 131K | ~175 | $0.40 | $0.40 |
| 37 | Llama 3.1 Nemotron 70B Instruct | NVIDIA | 131K | ~175 | $1.20 | $1.20 |
| 38 | Llama 3.1 Nemotron Ultra 253B v1 | NVIDIA | 131K | ~175 | $0.60 | $1.80 |
| 39 | Llama 3.2 11B Vision Instruct | Meta | 131K | ~175 | $0.05 | $0.05 |
| 40 | Llama 3.2 3B Instruct (free) | Meta | 131K | ~175 | Free | Free |
| 41 | Llama 3.3 70B Instruct | Meta | 131K | ~175 | $0.10 | $0.32 |
| 42 | Llama 3.3 Nemotron Super 49B V1.5 | NVIDIA | 131K | ~175 | $0.10 | $0.40 |
| 43 | Llama Guard 3 8B | Meta | 131K | ~175 | $0.02 | $0.06 |
| 44 | LongCat Flash Chat | Meituan | 131K | ~175 | $0.20 | $0.80 |
| 45 | Maestro Reasoning | arcee-ai | 131K | ~175 | $0.90 | $3.30 |
| 46 | Ministral 3 3B 2512 | Mistral AI | 131K | ~175 | $0.10 | $0.10 |
| 47 | Mistral Large 2407 | Mistral AI | 131K | ~175 | $2.00 | $6.00 |
| 48 | Mistral Large 2411 | Mistral AI | 131K | ~175 | $2.00 | $6.00 |
| 49 | Mistral Medium 3 | Mistral AI | 131K | ~175 | $0.40 | $2.00 |
| 50 | Mistral Medium 3.1 | Mistral AI | 131K | ~175 | $0.40 | $2.00 |
| 51 | Mistral Nemo | Mistral AI | 131K | ~175 | $0.02 | $0.04 |
| 52 | Mistral Small 3.1 24B | Mistral AI | 131K | ~175 | $0.03 | $0.11 |
| 53 | Nemotron Nano 12B 2 VL | NVIDIA | 131K | ~175 | $0.20 | $0.60 |
| 54 | Nemotron Nano 9B V2 | NVIDIA | 131K | ~175 | $0.04 | $0.16 |
| 55 | Pixtral Large 2411 | Mistral AI | 131K | ~175 | $2.00 | $6.00 |
| 56 | Qwen VL Max | Alibaba | 131K | ~175 | $0.52 | $2.08 |
| 57 | Qwen VL Plus | Alibaba | 131K | ~175 | $0.14 | $0.41 |
| 58 | Qwen-Turbo | Alibaba | 131K | ~175 | $0.03 | $0.13 |
| 59 | Qwen3 235B A22B | Alibaba | 131K | ~175 | $0.45 | $1.82 |
| 60 | Qwen3 235B A22B Thinking 2507 | Alibaba | 131K | ~175 | $0.15 | $1.50 |
| 61 | Qwen3 30B A3B Thinking 2507 | Alibaba | 131K | ~175 | $0.08 | $0.40 |
| 62 | Qwen3 Next 80B A3B Thinking | Alibaba | 131K | ~175 | $0.10 | $0.78 |
| 63 | Qwen3 VL 235B A22B Thinking | Alibaba | 131K | ~175 | $0.26 | $2.60 |
| 64 | Qwen3 VL 30B A3B Instruct | Alibaba | 131K | ~175 | $0.13 | $0.52 |
| 65 | Qwen3 VL 30B A3B Thinking | Alibaba | 131K | ~175 | $0.13 | $1.56 |
| 66 | Qwen3 VL 32B Instruct | Alibaba | 131K | ~175 | $0.10 | $0.42 |
| 67 | Qwen3 VL 8B Instruct | Alibaba | 131K | ~175 | $0.08 | $0.50 |
| 68 | Qwen3 VL 8B Thinking | Alibaba | 131K | ~175 | $0.12 | $1.36 |
| 69 | QwQ 32B | Alibaba | 131K | ~175 | $0.15 | $0.58 |
| 70 | R1 Distill Llama 70B | DeepSeek | 131K | ~175 | $0.70 | $0.80 |
| 71 | Spotlight | arcee-ai | 131K | ~175 | $0.18 | $0.18 |
| 72 | Tongyi DeepResearch 30B A3B | Alibaba | 131K | ~175 | $0.09 | $0.45 |
| 73 | Trinity Mini | arcee-ai | 131K | ~175 | $0.04 | $0.15 |
| 74 | Trinity Mini (free) | arcee-ai | 131K | ~175 | Free | Free |
| 75 | Virtuoso Large | arcee-ai | 131K | ~175 | $0.75 | $1.20 |
| 76 | Granite 4.0 Micro | IBM | 131K | ~175 | $0.02 | $0.11 |
| 77 | Kimi K2 0711 | Moonshot AI | 131K | ~175 | $0.55 | $2.20 |
| 78 | Trinity Large Preview (free) | arcee-ai | 131K | ~175 | Free | Free |
| 79 | Cogito v2.1 671B | deepcogito | 128K | ~171 | $1.25 | $1.25 |
| 80 | Command R (08-2024) | Cohere | 128K | ~171 | $0.15 | $0.60 |
| 81 | Command R+ (08-2024) | Cohere | 128K | ~171 | $2.50 | $10.00 |
| 82 | Command R7B (12-2024) | Cohere | 128K | ~171 | $0.04 | $0.15 |
| 83 | GPT Audio | OpenAI | 128K | ~171 | $2.50 | $10.00 |
| 84 | GPT Audio Mini | OpenAI | 128K | ~171 | $0.60 | $2.40 |
| 85 | GPT-4 Turbo | OpenAI | 128K | ~171 | $10.00 | $30.00 |
| 86 | GPT-4 Turbo (older v1106) | OpenAI | 128K | ~171 | $10.00 | $30.00 |
| 87 | GPT-4 Turbo Preview | OpenAI | 128K | ~171 | $10.00 | $30.00 |
| 88 | GPT-4o | OpenAI | 128K | ~171 | $2.50 | $10.00 |
| 89 | GPT-4o (2024-05-13) | OpenAI | 128K | ~171 | $5.00 | $15.00 |
| 90 | GPT-4o (2024-08-06) | OpenAI | 128K | ~171 | $2.50 | $10.00 |
| 91 | GPT-4o (2024-11-20) | OpenAI | 128K | ~171 | $2.50 | $10.00 |
| 92 | GPT-4o (extended) | OpenAI | 128K | ~171 | $6.00 | $18.00 |
| 93 | GPT-4o Audio | OpenAI | 128K | ~171 | $2.50 | $10.00 |
| 94 | GPT-4o Search Preview | OpenAI | 128K | ~171 | $2.50 | $10.00 |
| 95 | GPT-4o-mini | OpenAI | 128K | ~171 | $0.15 | $0.60 |
| 96 | GPT-4o-mini (2024-07-18) | OpenAI | 128K | ~171 | $0.15 | $0.60 |
| 97 | GPT-4o-mini Search Preview | OpenAI | 128K | ~171 | $0.15 | $0.60 |
| 98 | GPT-5 Chat | OpenAI | 128K | ~171 | $1.25 | $10.00 |
| 99 | GPT-5.1 Chat | OpenAI | 128K | ~171 | $1.25 | $10.00 |
| 100 | GPT-5.2 Chat | OpenAI | 128K | ~171 | $1.75 | $14.00 |
| 101 | GPT-5.3 Chat | OpenAI | 128K | ~171 | $1.75 | $14.00 |
| 102 | Mercury | Inception | 128K | ~171 | $0.25 | $0.75 |
| 103 | Mercury 2 | Inception | 128K | ~171 | $0.25 | $0.75 |
| 104 | Mercury Coder | Inception | 128K | ~171 | $0.25 | $0.75 |
| 105 | Mistral Large | Mistral AI | 128K | ~171 | $2.00 | $6.00 |
| 106 | Mistral Small 3.1 24B (free) | Mistral AI | 128K | ~171 | Free | Free |
| 107 | Mistral Small 3.2 24B | Mistral AI | 128K | ~171 | $0.07 | $0.20 |
| 108 | Nemotron Nano 12B 2 VL (free) | NVIDIA | 128K | ~171 | Free | Free |
| 109 | Nemotron Nano 9B V2 (free) | NVIDIA | 128K | ~171 | Free | Free |
| 110 | Nova Micro 1.0 | Amazon | 128K | ~171 | $0.04 | $0.14 |
| 111 | Olmo 2 32B Instruct | Allen AI | 128K | ~171 | $0.05 | $0.20 |
| 112 | Qwen2.5 VL 32B Instruct | Alibaba | 128K | ~171 | $0.20 | $0.60 |
| 113 | Solar Pro 3 | Upstage | 128K | ~171 | $0.15 | $0.60 |
| 114 | Sonar Deep Research | Perplexity | 128K | ~171 | $2.00 | $8.00 |
| 115 | Sonar Reasoning Pro | Perplexity | 128K | ~171 | $2.00 | $8.00 |
| 116 | UI-TARS 7B | ByteDance | 128K | ~171 | $0.10 | $0.20 |
Moderate context windows suitable for shorter documents, code files, and focused conversations.
| # | Model | Provider | Context Window | Pages | Input / 1M | Output / 1M |
|---|---|---|---|---|---|---|
| 1 | Sonar | Perplexity | 127K | ~169 | $1.00 | $1.00 |
| 2 | ERNIE 4.5 300B A47B | Baidu | 123K | ~164 | $0.28 | $1.10 |
| 3 | ERNIE 4.5 VL 424B A47B | Baidu | 123K | ~164 | $0.42 | $1.25 |
| 4 | ERNIE 4.5 21B A3B | Baidu | 120K | ~160 | $0.07 | $0.28 |
| 5 | Llama 3.2 3B Instruct | Meta | 80K | ~107 | $0.05 | $0.34 |
| 6 | Llama 3.3 70B Instruct (free) | Meta | 66K | ~87 | Free | Free |
| 7 | MiniMax M2-her | MiniMax | 66K | ~87 | $0.30 | $1.20 |
| 8 | Mixtral 8x22B Instruct | Mistral AI | 66K | ~87 | $2.00 | $6.00 |
| 9 | Nano Banana 2 (Gemini 3.1 Flash Image Preview) | 66K | ~87 | $0.50 | $3.00 | |
| 10 | Nano Banana Pro (Gemini 3 Pro Image Preview) | 66K | ~87 | $2.00 | $12.00 | |
| 11 | Olmo 3 32B Think | Allen AI | 66K | ~87 | $0.15 | $0.50 |
| 12 | Olmo 3.1 32B Instruct | Allen AI | 66K | ~87 | $0.20 | $0.60 |
| 13 | Olmo 3.1 32B Think | Allen AI | 66K | ~87 | $0.15 | $0.50 |
| 14 | WizardLM-2 8x22B | Microsoft | 66K | ~87 | $0.62 | $0.62 |
| 15 | R1 | DeepSeek | 64K | ~85 | $0.70 | $2.50 |
| 16 | Llama 3.2 1B Instruct | Meta | 60K | ~80 | $0.03 | $0.20 |
| 17 | Qwen3 14B | Alibaba | 41K | ~55 | $0.06 | $0.24 |
| 18 | Qwen3 30B A3B | Alibaba | 41K | ~55 | $0.08 | $0.28 |
| 19 | Qwen3 32B | Alibaba | 41K | ~55 | $0.08 | $0.24 |
| 20 | Qwen3 4B (free) | Alibaba | 41K | ~55 | Free | Free |
| 21 | Qwen3 8B | Alibaba | 41K | ~55 | $0.05 | $0.40 |
| 22 | Coder Large | arcee-ai | 33K | ~44 | $0.50 | $0.80 |
| 23 | DeepSeek V3.1 | DeepSeek | 33K | ~44 | $0.15 | $0.75 |
| 24 | Gemma 3 12B (free) | 33K | ~44 | Free | Free | |
| 25 | Gemma 3 4B (free) | 33K | ~44 | Free | Free | |
| 26 | Gemma 3n 4B | 33K | ~44 | $0.02 | $0.04 | |
| 27 | LFM2-2.6B | Liquid AI | 33K | ~44 | $0.01 | $0.02 |
| 28 | LFM2-24B-A2B | Liquid AI | 33K | ~44 | $0.03 | $0.12 |
| 29 | LFM2-8B-A1B | Liquid AI | 33K | ~44 | $0.01 | $0.02 |
| 30 | LFM2.5-1.2B-Instruct (free) | Liquid AI | 33K | ~44 | Free | Free |
| 31 | LFM2.5-1.2B-Thinking (free) | Liquid AI | 33K | ~44 | Free | Free |
| 32 | Llama 3.1 405B (base) | Meta | 33K | ~44 | $4.00 | $4.00 |
| 33 | Mistral Small 3 | Mistral AI | 33K | ~44 | $0.05 | $0.08 |
| 34 | Mistral Small Creative | Mistral AI | 33K | ~44 | $0.10 | $0.30 |
| 35 | Mixtral 8x7B Instruct | Mistral AI | 33K | ~44 | $0.54 | $0.54 |
| 36 | Nano Banana (Gemini 2.5 Flash Image) | 33K | ~44 | $0.30 | $2.50 | |
| 37 | Pixtral 12B | Mistral AI | 33K | ~44 | $0.10 | $0.10 |
| 38 | Qwen-Max | Alibaba | 33K | ~44 | $1.04 | $4.16 |
| 39 | Qwen2.5 72B Instruct | Alibaba | 33K | ~44 | $0.12 | $0.39 |
| 40 | Qwen2.5 7B Instruct | Alibaba | 33K | ~44 | $0.04 | $0.10 |
| 41 | Qwen2.5 Coder 32B Instruct | Alibaba | 33K | ~44 | $0.66 | $1.00 |
| 42 | Qwen2.5 Coder 7B Instruct | Alibaba | 33K | ~44 | $0.03 | $0.09 |
| 43 | Qwen2.5 VL 72B Instruct | Alibaba | 33K | ~44 | $0.80 | $0.80 |
| 44 | Qwen2.5-VL 7B Instruct | Alibaba | 33K | ~44 | $0.20 | $0.20 |
| 45 | R1 Distill Qwen 32B | DeepSeek | 33K | ~44 | $0.29 | $0.29 |
| 46 | Rnj 1 Instruct | essentialai | 33K | ~44 | $0.15 | $0.15 |
| 47 | Saba | Mistral AI | 33K | ~44 | $0.20 | $0.60 |
| 48 | Voxtral Small 24B 2507 | Mistral AI | 32K | ~43 | $0.10 | $0.30 |
A model's context window is the total number of tokens (roughly words) it can process in a single request. This includes both your input prompt and the model's output. A model with a 128K context window can process about 170 pages of text at once, while a 1M-token model can handle roughly 1,300 pages -- enough for entire books or large codebases.
With small context windows (under 32K), you must chunk documents and use retrieval-augmented generation (RAG). Large context models eliminate this complexity for many workloads: analyzing full legal contracts, reviewing entire repositories, summarizing research paper collections, or maintaining very long conversations with full history retained.
Not all context is created equal. Some models perform well on "needle in a haystack" tests at their full context length, while others degrade on information retrieval when prompts get very long. The advertised context window is the maximum, but effective performance may vary. Check our leaderboard for quality scores that account for real-world performance.
Using a large context window means sending more tokens per request, which increases cost. For example, filling a 1M-token context at $3/1M input tokens costs $3 per request. For cost-sensitive workloads, consider whether RAG with a smaller context model might be more efficient than filling a large context window end to end.
Dive deeper into model capabilities, compare context windows side by side, or see overall rankings across all dimensions.
A large context window model can process long inputs - from 128K tokens (about 100 pages) up to 2M tokens (about 1,500 pages). This allows analyzing entire codebases, long documents, or extensive conversation histories in a single request.
Large context models excel when you need the AI to reason across an entire document at once. RAG (Retrieval-Augmented Generation) is better for searching across massive document collections. Large context is simpler to implement but costs more per request.
No. While many models advertise 1M+ token context windows, their effective recall can vary significantly. Some models lose accuracy when important information is buried in the middle of very long contexts. Our rankings account for practical performance, not just advertised limits.