Rankings

The best models, ranked.

Frontier models ranked by benchmark intelligence — reasoning, coding, math and knowledge evals — with real-world request volume shown alongside. Sort by quality to find the strongest model, or by requests to see what the market actually runs.

40 models · benchmarks 2026-09-06 · usage 2026-07-10

Top models

#ModelProviderIntelligenceCodingRequestsShare
1Claude Fable 5.1
anthropic/claude-fable-5.1
Anthropic56.881.6
2Claude Opus 5
anthropic/claude-opus-5
Anthropic54.178.0
3Claude Fable 5
anthropic/claude-fable-5
Anthropic53.276.52420%
4GPT-5.6 Sol
openai/gpt-5.6-sol
OpenAI51.377.4
5Grok 4.6
xai/grok-4.6
xAI50.676.8
6Kimi K3
moonshot/kimi-k3
Moonshot50.276.2
7Glm-5.3
zai-org/glm-5.3
Zai-org48.674.8
8Gemini 3.8 Flash
google/gemini-3.8-flash
Google47.176.3
9Qwen3.8 Max
alibaba/qwen3.8-max
Alibaba46.971.8
10GPT-5.6 Terra
openai/gpt-5.6-terra
OpenAI46.876.7
11Claude Opus 4.8
anthropic/claude-opus-4.8
Anthropic46.474.3370%
12Glm-5.3-Flash
zai-org/glm-5.3-flash
Zai-org46.271.5
13GPT-5.5
openai/gpt-5.5
OpenAI45.674.935.2K3.1%
14Grok 4.5
xai/grok-4.5
xAI45.572.4
15Gemini 3.7 Flash
google/gemini-3.7-flash
Google45.276.1
16Claude Sonnet 5
anthropic/claude-sonnet-5
Anthropic45.171.5
17Claude Opus 4.7
anthropic/claude-opus-4.7
Anthropic44.373.610%
18GPT-5.6 Luna
openai/gpt-5.6-luna
OpenAI43.471.4
19GPT-5.4
openai/gpt-5.4
OpenAI42.871.119.2K1.7%
20GLM-5.2
zai-org/glm-5.2
Zai-org42.568.816.1K1.4%
21DeepSeek V4 Pro
deepseek/deepseek-v4-pro
Deepseek42.168.857.9K5%
22Qwen3.8-27b
qwen/qwen3.8-27b
Alibaba Qwen41.468.1
23Gemini 3.6 Flash
google/gemini-3.6-flash
Google40.369.2
24Gemini 3.5 Flash
google/gemini-3.5-flash
Google39.770.1
25Claude Sonnet 4.6
anthropic/claude-sonnet-4.6
Anthropic38.563.0
26Qwen3.7 Max
alibaba/qwen3.7-max
Alibaba37.066.0
27Gemini 3.1 Pro
google/gemini-3.1-pro
Google36.768.840.7K3.5%
28Claude Opus 4.6
anthropic/claude-opus-4.6
Anthropic36.4
29Kimi-K2.6
moonshot/kimi-k2.6
Moonshot35.861.89.4K0.8%
30MiniMax M3
minimax/m3
MiniMax35.758.6
31Kimi-K2.7-Code
moonshot/kimi-k2.7-code
Moonshot33.960.84.3K0.4%
32Claude Opus 4.5
anthropic/claude-opus-4.5
Anthropic33.610%
33GPT-5.4 Mini
openai/gpt-5.4-mini
OpenAI31.956.160.6K5.3%
34GPT-5.4 Nano
openai/gpt-5.4-nano
OpenAI30.956.114.7K1.3%
35Qwen3.7 Plus
alibaba/qwen3.7-plus
Alibaba30.855.9
36Gemini 3 Flash
google/gemini-3-flash
Google30.7248.8K21.6%
37MiniMax M2.7
minimax/m2.7
MiniMax30.152.6
38Grok 4.20 Non-Reasoning
xai/grok-4.20-0309-non-reasoning
xAI29.5
39Grok 4.20 Reasoning
xai/grok-4.20-0309-reasoning
xAI29.5
40Grok 4.3
xai/grok-4.3
xAI29.342.234.4K3%

By provider · demand

ProviderModelsRequests
Google13766.6K
OpenAI27246.4K
Deepseek257.9K
xAI635.0K
Zai-org416.1K
Moonshot313.6K
Anthropic123.7K
Mistral12.6K
Ibm-granite12.4K
Meta51.7K
Black Forest Labs91.1K
Inworld2947
How this is built

Benchmarks for quality, usage for demand.

Intelligence and coding scores come from Artificial Analysis evals baked into the catalog on every release; request volume is real aggregate demand across the open model ecosystem, refreshed daily. For the full benchmark + pricing table, see the model leaderboard.