Exact score

#	Model	Arena	Acc Δ	Exact	ROI Δ	N
1	Claude Opus 5 Anthropic	71.5	+2.8%	14%	+9.8%	316
2	DeepSeek V4 Pro DeepSeek	51.6	−1.2%	13%	+2.8%	443
3	Grok 4.5 xAI	58.2	0.0%	13%	+5.0%	340
4	Mistral Large 3 Mistral	57.3	−0.2%	12%	+8.3%	433
5	MiniMax M3 MiniMax	55.0	0.0%	12%	+3.9%	340
6	GPT-5.6 Sol OpenAI	55.6	+0.1%	12%	+4.2%	339
7	Qwen3.7 Plus Alibaba	54.9	0.0%	12%	+3.8%	340
8	Gemini 3.7 Flash Google	50.6	−0.8%	12%	+1.1%	239
9	MiMo v2.5-Pro Xiaomi	46.9	−1.2%	11%	+2.3%	444
10	Kimi K3 Moonshot	56.2	+1.2%	11%	+4.8%	324
11	Muse Spark 1.2 Meta	46.7	−1.6%	11%	+1.6%	259
12	GLM-5.2 Z.ai	63.5	+2.1%	10%	+12.8%	339
13	Nemotron 3 Ultra NVIDIA	48.7	−1.8%	10%	+6.0%	336

1Claude Opus 5 Anthropic14%Exact

Arena

71.5

Acc Δ

+2.8%

Exact

14%

ROI Δ

+9.8%

2DeepSeek V4 Pro DeepSeek13%Exact

Arena

51.6

Acc Δ

−1.2%

Exact

13%

ROI Δ

+2.8%

3Grok 4.5 xAI13%Exact

Arena

58.2

Acc Δ

0.0%

Exact

13%

ROI Δ

+5.0%

4Mistral Large 3 Mistral12%Exact

Arena

57.3

Acc Δ

−0.2%

Exact

12%

ROI Δ

+8.3%

5MiniMax M3 MiniMax12%Exact

Arena

55.0

Acc Δ

0.0%

Exact

12%

ROI Δ

+3.9%

6GPT-5.6 Sol OpenAI12%Exact

Arena

55.6

Acc Δ

+0.1%

Exact

12%

ROI Δ

+4.2%

7Qwen3.7 Plus Alibaba12%Exact

Arena

54.9

Acc Δ

0.0%

Exact

12%

ROI Δ

+3.8%

8Gemini 3.7 Flash Google12%Exact

Arena

50.6

Acc Δ

−0.8%

Exact

12%

ROI Δ

+1.1%

9MiMo v2.5-Pro Xiaomi11%Exact

Arena

46.9

Acc Δ

−1.2%

Exact

11%

ROI Δ

+2.3%

10Kimi K3 Moonshot11%Exact

Arena

56.2

Acc Δ

+1.2%

Exact

11%

ROI Δ

+4.8%

11Muse Spark 1.2 Meta11%Exact

Arena

46.7

Acc Δ

−1.6%

Exact

11%

ROI Δ

+1.6%

12GLM-5.2 Z.ai10%Exact

Arena

63.5

Acc Δ

+2.1%

Exact

10%

ROI Δ

+12.8%

13Nemotron 3 Ultra NVIDIA10%Exact

Arena

48.7

Acc Δ

−1.8%

Exact

10%

ROI Δ

+6.0%

Every active model is scored on identical terms, so the ranking compares like with like. How scoring works →