Flagship ranking · 13 models

Overall LLM Leaderboard

A weighted composite ranking across seven dimensions — intelligence (30%), coding (20%), reasoning (15%), Chinese (10%), long context (8%), multimodal (7%) and value (10%). Click any model for the full breakdown.

Composite scores are edited estimates from public sources, not official results. See how we calculate them. Data as of 2026-11-20.
# Model OverallIntelligenceCodingReasoningChineseContextPrice (in/out)
1 Claude Opus 5.5
Anthropic · US
93.3989997881M$4 / $20
2 GPT-6 Astra
OpenAI · US
90.1959196861.1M$10 / $50
3 Kimi K3.1 Pro
Moonshot AI · CN
89.8919390921M$2 / $10
4 Gemini 4 Argon
Google · US
89.7928688901M$2 / $10
5 GPT-6.1 Sol
OpenAI · US
89.3908889851.1M$2 / $10
6 Claude Sonnet 5.5
Anthropic · US
88.9899287851M$2 / $10
7 Qwen3.8 Max
Alibaba Cloud · CN
88.8888687981M$2 / $6
8 GLM-5.3
Z.ai (Zhipu) · CN
84.584898391200K$0.60 / $2.20
9 Grok 4.7
xAI · US
84.386878478500K$2 / $6
10 DeepSeek V4.1 Flash
DeepSeek · CN
83.6828880881M$0.15 / $0.60
11 Doubao Seed 2.1 Pro
ByteDance · CN
82.781788095256K$0.30 / $1.20
12 Mistral Large 4
Mistral AI · EU
80.5808079761M$0.70 / $2.10
13 Llama 4 Maverick
Meta · US
78.5777675741M$0.20 / $0.60

Price shown as USD per 1M input / output tokens. Context is the maximum window.