Skip to content
AIDC热点 LEADERBOARD

AI model rankings

Real benchmarks across many capabilities, to find the stronger all-round model.

24 benchmarks·12 organizations·Updated 11/02 14:50

Overall rankingTOP 30

Ranked by shared evidence from multiple public benchmarks
Top 30 models, sortable by column
Model
01Claude Opus 5.5AnthropicLive 2026-09-17 · AccumulatingCached input ¥1.33Input ¥26.61 · Output ¥133.0698.5
02Kimi K3.1 ProMoonshot AILive 2026-10-23 · Evidence-sensitiveCached input ¥1.3Input ¥13 · Output ¥6898.4
03GPT-6 AstraOpenAILive 2026-09-03 · Evidence-sensitiveCached input ¥6.65Input ¥66.53 · Output ¥332.6498.1
04Gemini 4 ArgonGoogleLive 2026-09-30 · Evidence-sensitiveCached input ¥0.67Input ¥13.31 · Output ¥66.5397.4
05Claude Fable 5.1AnthropicLive 2026-09-01 · Evidence-sensitiveCached input ¥1.66Input ¥66.53 · Output ¥332.6497.2
06Muse WonderMetaLive — · Evidence-sensitiveCached input ¥1.33Input ¥26.61 · Output ¥133.0696.9
07GPT-6.1 SolOpenAILive — · Evidence-sensitiveCached input ¥0.67Input ¥13.31 · Output ¥66.5396.1
08Claude Opus 5AnthropicLive 2026-07-24 · Evidence-sensitiveCached input ¥3.33Input ¥33.26 · Output ¥166.3294.9
09GPT-6.1 TerraOpenAILive 2026-10-26 · Evidence-sensitiveCached input —Input — · Output —94.0
10Claude Fable 5AnthropicLive 2026-06-09 · Evidence-sensitiveCached input ¥6.65Input ¥66.53 · Output ¥332.6493.9
11Claude Sonnet 5.5AnthropicLive — · Evidence-sensitiveCached input ¥0.67Input ¥13.31 · Output ¥66.5393.8
12muse-wonder-1.0MetaLive — · Evidence-sensitiveCached input ¥1.33Input ¥26.61 · Output ¥133.0693.7
13Muse Spark 1.3MetaLive 2026-09-02 · Evidence-sensitiveCached input —Input ¥8.32 · Output ¥28.2793.5
14GPT-6 SolOpenAILive 2026-09-22 · Evidence-sensitiveCached input ¥1.33Input ¥13.31 · Output ¥66.5391.8
15GPT-5.6 SolOpenAILive 2026-07-09 · Evidence-sensitiveCached input ¥2.66Input ¥26.61 · Output ¥133.0691.4
16Grok 4.7xAILive 2026-09-21 · Evidence-sensitiveCached input ¥3.33Input ¥13.31 · Output ¥39.9288.8
17MiMo V 2.6 ProXiaomiLive 2026-09-21 · Evidence-sensitiveCached input —Input ¥3 · Output ¥687.7
18Qwen3.8 MaxAlibabaLive 2026-07-19 · Evidence-sensitiveCached input —Input ¥12 · Output ¥3687.1
19deepseek-v4.1-flash-1017DeepSeekLive — · Evidence-sensitiveCached input —Input — · Output —85.9
20GLM-5.3Z.aiLive 2026-08-18 · Evidence-sensitiveCached input ¥1.73Input ¥9.31 · Output ¥29.2783.6
21Grok 4.6xAILive 2026-08-12 · Evidence-sensitiveCached input ¥3.33Input ¥13.31 · Output ¥39.9282.1
22Step 5 PreviewStepFunLive 2026-09-18 · Evidence-sensitiveCached input ¥0.33Input ¥6.65 · Output ¥17.9679.7
23Kimi K3Moonshot AILive 2026-07-16 · Evidence-sensitiveCached input ¥2Input ¥20 · Output ¥10079.0
24Claude Haiku 5.5AnthropicLive — · Evidence-sensitiveCached input ¥0.0665Input ¥0.67 · Output ¥3.3376.8
25GPT-5.6 TerraOpenAILive 2026-07-09 · Evidence-sensitiveCached input ¥1.33Input ¥13.31 · Output ¥79.8376.5
26GPT-5.4 2026-03-05OpenAILive 2026-03-05 · Evidence-sensitiveCached input ¥1.66Input ¥16.63 · Output ¥99.7976.5
27Claude Opus 4.8AnthropicLive 2026-05-28 · Evidence-sensitiveCached input ¥3.33Input ¥33.26 · Output ¥166.3276.5
28Gemini 3.8 FlashGoogleLive 2026-09-02 · Evidence-sensitiveCached input ¥0.5Input ¥4.99 · Output ¥24.9576.0
29GLM-5.3 FlashZ.aiLive 2026-08-26 · Evidence-sensitiveCached input ¥0.23Input ¥0.8 · Output ¥2.875.4
30Claude Opus 4.7AnthropicLive 2026-04-16 · Evidence-sensitiveCached input ¥3.33Input ¥33.26 · Output ¥166.3274.8

Each board shows up to 30 models

The consensus index is not an accuracy rate; tied scores still show ranks set by shared evidence.

How to read this board

综合多家公开评测,不同模型的参评覆盖不同。Price does not affect rank, missing benchmarks score no zero, and the index is not an accuracy rate.

How it's calculated →

About pricing

API prices come from vendors' official sites, shown per million tokens.USD prices are converted to CNY at a rate of 2026-10-30.Cached price means input price on a cache hit; cache writes, storage and subscription fees are separate.