MMLU-Pro leaderboard

Harder, reasoning-focused multiple choice across 14 subjects, from law to engineering. Claude Fable 5.1 leads with 92.38%.

Models
120
Leader
Claude Fable 5.1
Source
Vals AI
As of
Oct 24, 2026
  1. 1Claude Fable 5.192.4%
  2. 2Claude Opus 591.6%
  3. 3Claude Fable 591.5%
  4. 4Gemini 3.1 Pro Preview91.0%
  5. 5Gemini 3.8 Flash90.2%
  6. 6Gemini 3.7 Flash90.1%
  7. 7Gemini 3 Pro90.1%
  8. 8Claude Opus 4.789.9%
  9. 9Claude Opus 4.889.6%
  10. 10Gemini 3.5 Flash89.5%
MMLU-Pro, accuracy against cost per test
Open weightsProprietaryBest for the price
20%40%60%80%100%$.1$1$10Cost per test, log scale →Claude Fable 5.1Gemini 3.1 Pro PreviewGemini 3 FlashGLM-5.3-FlashGPT-5 nano

Full leaderboard

RankModelLabAccuracyCost / testHarnessPrice in / out
1Claude Fable 5.1Anthropic92.38%$10 / $50—$10 / $50
2Claude Opus 5Anthropic91.59%$5 / $25—$5 / $25
3Claude Fable 5Anthropic91.50%$10 / $50—$10 / $50
4Gemini 3.1 Pro PreviewGoogle90.99%$2 / $12—$2 / $12
5Gemini 3.8 FlashGoogle90.22%$1.5 / $7.5—$0.82 / $4.13
6Gemini 3.7 FlashGoogle90.12%$1.5 / $7.5—$0.75 / $3.75
7Gemini 3 ProGoogle90.10%$2 / $12——
8Claude Opus 4.7Anthropic89.87%$5 / $25—$5 / $25
9Claude Opus 4.8Anthropic89.58%$5 / $25—$5 / $25
10Gemini 3.5 FlashGoogle89.52%$1.5 / $9—$1.50 / $9
11Grok 4.6xAI89.40%$2 / $6—$2 / $6
12Qwen3.7-Max-PreviewAlibaba89.31%$2.5 / $7.5—$1.48 / $4.42
13Gemini 3.6 FlashGoogle89.28%$1.5 / $7.5—$0.75 / $3.75
14Grok 4.5xAI89.22%$2 / $6—$2 / $6
15Claude Opus 4.6Anthropic89.11%$5 / $25—$5 / $25
16GPT-5.6 SolOpenAI89.10%$4 / $20—$4 / $20
17Muse Spark 1.1Meta88.73%$1.25 / $4.25—$1.25 / $4.25
18Qwen3.8-MaxAlibaba88.60%$2 / $6—$2 / $6
19Gemini 3 FlashGoogle88.59%$0.5 / $3—$0.50 / $3
20Muse Spark 1.2Meta88.28%$1.25 / $4.25—$1.25 / $4.25
21GPT-5.5OpenAI88.14%$5 / $30—$5 / $30
22Kimi K3Moonshot AI87.97%$3 / $15—$0.24 / $11
24Qwen3.6-PlusAlibaba87.67%$0.5 / $3—$0.33 / $1.95
25Kimi K2.6Moonshot AI87.57%$0.95 / $4—$0.65 / $3.41
26Claude Sonnet 5Anthropic87.55%$2 / $10—$2 / $10
27GPT-5.4OpenAI87.48%$2.5 / $15—$2.50 / $15
28Claude Sonnet 4.5Anthropic87.36%$3 / $15—$3 / $15
29Claude Sonnet 4.6Anthropic87.34%$3 / $15—$3 / $15
30Muse SparkMeta87.32%N/A——
33Claude Opus 4.1Anthropic87.21%$15 / $75—$15 / $75
34Qwen 3.5 PlusAlibaba87.18%$0.4 / $2.4—$0.30 / $1.80
35MiniMax-M2.1 PreviewMiniMax87.05%$0.3 / $1.2—$0.30 / $1.20
36DeepSeek-V4-ProDeepSeek86.97%$1.32 / $3.96—$0.66 / $1.98
37GLM-5.1Z.ai86.90%$1 / $3.2—$1.40 / $4.40
38GLM-5.3Z.ai86.77%$1.4 / $4.4—$0.079 / $1.40
39GLM-5.2Z.ai86.71%$1.4 / $4.4—$0.034 / $7
40GPT-5.6 TerraOpenAI86.66%$2 / $12—$2 / $12
41GPT-5OpenAI86.54%$1.25 / $10—$1.25 / $10
42GPT-5.1OpenAI86.38%$1.25 / $10—$1.25 / $10
43InklingThinky86.30%$1 / $4.05—$1 / $4.05
44Grok 4.20xAI86.25%$2 / $6—$1.25 / $2.50
45Gemini 3.1 Flash-Lite PreviewGoogle86.24%$0.25 / $1.5—$0.25 / $1.50
46GPT-5.2OpenAI86.23%$1.75 / $14—$1.75 / $14
47DeepSeek-V4-FlashDeepSeek86.21%$0.44 / $1.32—$0.042 / $0.18
48Claude Opus 4Anthropic86.17%$15 / $75——
49GLM-5.3-FlashZ.ai86.06%$0.075 / $0.25—$0.15 / $0.50
50GPT-5.6 LunaOpenAI86.04%$0.2 / $1.2—$0.20 / $1.20
51GLM-5Z.ai86.03%$1 / $3.2—$1 / $3.20
52Kimi K2.5Moonshot AI85.91%$0.6 / $3—$0.50 / $2.50
53Grok 4.3xAI85.84%$1.25 / $2.5—$1.25 / $2.50
54Gemini 3.5 Flash-LiteGoogle85.84%$0.3 / $2.5—$0.30 / $2.50
55Nemotron 3 UltraNVIDIA85.76%N/A—$0.60 / $2.40
56o3OpenAI85.59%$2 / $8—$2 / $8
57Claude Opus 4.5Anthropic85.59%$5 / $25—$5 / $25
58Inkling-SmallThinky85.57%$0.3 / $1.2—$0.45 / $1.20
59Grok 4xAI85.30%$3 / $15——
62MiMo-V2.5-ProXiaomi84.59%$0.435 / $0.87—$0.30 / $0.61
63GPT-5.4 miniOpenAI84.55%$0.75 / $4.5—$0.75 / $4.50
65Qwen3.8-27BAlibaba84.34%$0.5 / $3—$0.20 / $2.50
66MiniMax-M3MiniMax84.22%$0.6 / $2.4—$0.30 / $1.20
67Grok 4.1 FastxAI84.18%$0.2 / $0.5——
68Qwen3.5-FlashAlibaba84.06%$0.1 / $0.4—$0.065 / $0.26
69Gemini 2.5 ProGoogle84.06%$1.25 / $10—$1.25 / $10
73Qwen3-Max-PreviewAlibaba83.54%$1.2 / $6——
74o1OpenAI83.49%$15 / $60—$15 / $60
75DeepSeek-R1DeepSeek83.18%$1.35 / $5.4—$0.50 / $2.15
76DeepSeek-V3.2-ExpDeepSeek83.06%$0.56 / $1.68—$0.28 / $0.42
77MiMo-V2.5Xiaomi82.93%$0.14 / $0.28—$0.12 / $0.24
78GLM-4.7Z.ai82.74%$0.6 / $2.2—$0.54 / $1.98
80GPT-5 miniOpenAI82.23%$0.25 / $2—$0.25 / $2
81GLM-4.6Z.ai82.20%$0.6 / $2.2—$0.55 / $2.20
82Ling 3.0 FlashinclusionAI82.01%$0.075 / $0.22—$0.021 / $0.063
84Qwen3-235B-A22BAlibaba81.25%$0.22 / $0.88—$0.09 / $0.55
85GLM-4.5Z.ai81.22%$0.6 / $2.2—$0.60 / $2.20
86Kimi K2 ThinkingMoonshot AI81.07%$0.6 / $2.5—$0.60 / $2.50
87Claude 3.7 SonnetAnthropic80.66%$3 / $15——
88o4-miniOpenAI80.56%$1.1 / $4.4—$1.10 / $4.40
89GPT-4.1OpenAI80.50%$2 / $8—$2 / $8
90MiniMax-M2.7MiniMax80.43%$0.3 / $1.2—$0.30 / $1.20
91MiniMax-M2.5MiniMax80.09%$0.3 / $1.2—$0.30 / $1.20
92Grok 3 Mini ReasoningxAI80.01%$0.3 / $0.5——
93Grok 3 PreviewxAI79.95%$3 / $15——
97Claude Sonnet 4Anthropic79.43%$3 / $15——
98Llama 4 MaverickFireworks AI79.42%$0.22 / $0.88—$0.19 / $0.65
99Kimi K2 InstructMoonshot AI79.39%$1 / $3——
100gpt-oss-120bOpenAI79.17%$0.15 / $0.6—$0.15 / $0.60
102Claude Haiku 4.5Anthropic78.72%$1 / $5—$1 / $5
103o3-miniOpenAI78.69%$1.1 / $4.4—$1.10 / $4.40
104Gemini 2.5 Flash PreviewGoogle78.64%$0.1 / $0.4——
105Claude 3.5 SonnetAnthropic78.40%$3 / $15——
106Gemini 2.0 FlashGoogle77.38%$0.1 / $0.4——
107GPT-4.1 miniOpenAI77.22%$0.4 / $1.6—$0.40 / $1.60
108GPT-5.4 nanoOpenAI77.17%$0.2 / $1.25—$0.20 / $1.25
109GPT-5 nanoOpenAI76.07%$0.05 / $0.4—$0.05 / $0.40
110Grok 2xAI75.47%$2 / $10——
111Mistral Medium 3.5Mistral AI75.33%$1.5 / $7.5—$1.50 / $7.50
112Gemini 1.5 ProGoogle75.29%$1.25 / $5——
113Mistral MediumMistral AI75.29%$0.4 / $2——
114Grok 4.1 Fast Non-ReasoningxAI75.21%$0.2 / $0.5——
116DeepSeek-V3DeepSeek73.82%$0.9 / $0.9—$0.29 / $1.14
117GPT-4oOpenAI72.56%$2.5 / $10—$2.50 / $10
118gpt-oss-20bOpenAI71.64%$0.07 / $0.3—$0.018 / $0.09
120Grok 4 FastxAI70.34%$0.2 / $0.5——
121Llama 3.3 Instruct TurboMeta69.86%$0.88 / $0.88——
122Mistral LargeMistral AI69.71%$2 / $6—$0.50 / $1.50
123Llama 4 ScoutMeta69.63%$0.18 / $0.59—$0.10 / $0.30
124Llama 3.3 Nemotron SuperMeta69.58%N/A——
125Command ACohere69.17%$2.5 / $10—$0.30 / $1.50
126Laguna XS.2Poolside69.05%N/A——
127Laguna M.1Poolside68.84%N/A——
128Magistral MediumMistral AI68.66%$2 / $5——
130Gemini 1.5 FlashGoogle65.61%$0.075 / $0.3——
131Mistral SmallMistral AI64.44%$0.2 / $0.6——
132Claude 3.5 HaikuAnthropic64.12%$0.8 / $4——
133GPT-4.1 nanoOpenAI63.48%$0.1 / $0.4—$0.10 / $0.40
134GPT-4o miniOpenAI62.73%$0.15 / $0.6—$0.15 / $0.60
135Magistral Small 1.2Mistral AI62.13%$0.5 / $1.5——
136Jamba 1.6 LargeAI21 Labs49.78%$2 / $8——
137Command R+Cohere44.00%$2.5 / $10—$0.15 / $0.60
138Jamba 1.6 MiniAI21 Labs30.28%$0.2 / $0.4——

Related leaderboards

Sources

  • Vals AI, as of Oct 24, 2026
  • Provider list prices, as of Oct 24, 2026

Every number on this page comes from these sources; models without a published number show none. Download the data. Cite as “Nexus Explorer, nexus.inc/models”.