benchgap
General

BlenderWorld (text-only) leaderboard

As of 2026-11-18, the highest measured score on BlenderWorld (text-only) is 86.1% by Claude Opus 5.5. 98 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1GPT-6.2 Astra86.6%estimated ± 2.8 pp, low confidence
2Claude Fable 5.286.5%estimated ± 2.8 pp, low confidence
3GPT-6 Astra86.3%estimated ± 2.8 pp, low confidence
4Claude Opus 5.586.1%measured
5GPT-5.5 Pro85.7%estimated ± 4.0 pp, medium confidence
6GPT-5.4 Pro85.7%estimated ± 4.0 pp, medium confidence
7GPT-6.1 Terra85.7%estimated ± 2.8 pp, medium confidence
8Gemini 4 Carbon85.6%estimated ± 2.8 pp, medium confidence
9GPT-6.1 Sol85.3%measured
10Claude Sonnet 5.584.4%estimated ± 2.8 pp, medium confidence
11Gemini 4 Argon84.2%estimated ± 2.8 pp, medium confidence
12Gemini 3 Pro83.6%estimated ± 4.3 pp, medium confidence
13Claude Opus 582.9%estimated ± 2.8 pp, medium confidence
14Claude Opus 4.6 (Adaptive)82.8%estimated ± 4.3 pp, medium confidence
15Claude Fable 5.182.6%estimated ± 2.8 pp, medium confidence
16GPT-5.6 Sol82.6%estimated ± 2.8 pp, medium confidence
17GPT-5.581.9%estimated ± 2.8 pp, medium confidence
18GPT-6 Sol81.9%estimated ± 2.8 pp, medium confidence
19Grok 4.20 Multi-agent81.8%estimated ± 4.3 pp, medium confidence
20Grok 4.1 Fast (Reasoning)80.7%estimated ± 4.3 pp, medium confidence
21Claude Fable 580.2%estimated ± 2.8 pp, medium confidence
22Muse Wonder79.7%estimated ± 2.8 pp, medium confidence
23DeepSeek V4.1 Flash 101779.2%measured
24ERNIE 5.5 Flash Preview79.0%measured
25Gemini 3.8 Flash77.7%estimated ± 2.8 pp, medium confidence
26Hy476.8%measured
27Kimi K3.1 Pro75.0%estimated ± 2.8 pp, medium confidence
28Muse Spark Open74.3%estimated ± 4.3 pp, medium confidence
29Qwen3.8 Max73.2%estimated ± 2.8 pp, medium confidence
30Gemini 3.7 Flash72.8%estimated ± 2.8 pp, medium confidence
31GPT-5.472.8%estimated ± 2.8 pp, medium confidence
32Step 5 Preview72.5%estimated ± 4.3 pp, medium confidence
33Claude Opus 4.872.4%estimated ± 2.8 pp, medium confidence
34Claude Sonnet 572.1%estimated ± 2.8 pp, medium confidence
35GPT-5.6 Terra72.1%estimated ± 2.8 pp, medium confidence
36DeepSeek V4 Flash 073171.8%estimated ± 2.8 pp, medium confidence
37Gemini 3.1 Pro71.8%estimated ± 2.8 pp, medium confidence
38Grok 4.671.8%estimated ± 2.8 pp, medium confidence
39GLM-5.3-Flash71.4%measured
40Gemini 3.5 Flash71.3%estimated ± 2.8 pp, medium confidence
41Qwen3.7 Max71.3%estimated ± 2.8 pp, medium confidence
42Claude Haiku 5.570.9%estimated ± 2.8 pp, medium confidence
43Gemini 3.6 Flash70.9%estimated ± 2.8 pp, medium confidence
44Grok 4.770.7%estimated ± 2.8 pp, medium confidence
45o370.7%estimated ± 2.8 pp, medium confidence
46Claude Opus 4.770.6%estimated ± 2.8 pp, medium confidence
47Grok 4.570.6%estimated ± 4.0 pp, medium confidence
48Muse Spark 1.170.6%estimated ± 4.0 pp, medium confidence
49Grok 4.2070.6%estimated ± 4.0 pp, medium confidence
50Claude 3 Opus70.6%estimated ± 4.0 pp, low confidence
51Claude Haiku 4.570.6%estimated ± 4.0 pp, low confidence
52GLM-4.770.6%estimated ± 4.0 pp, medium confidence
53GLM-5.170.6%estimated ± 4.0 pp, low confidence
54GPT-4.170.6%estimated ± 4.0 pp, low confidence
55GPT-4.1 nano70.6%estimated ± 4.0 pp, low confidence
56GPT-4o70.6%estimated ± 4.0 pp, low confidence
57Grok 4.370.6%estimated ± 4.0 pp, low confidence
58Inkling70.6%estimated ± 4.0 pp, medium confidence
59Kimi K2.7 Code70.6%estimated ± 4.0 pp, low confidence
60Muse Spark 1.270.6%estimated ± 4.0 pp, medium confidence
61o170.6%estimated ± 4.0 pp, low confidence
62Seed 2.1 Pro70.6%estimated ± 4.0 pp, medium confidence
63Seed 2.1 Turbo70.6%estimated ± 4.0 pp, medium confidence
64Kimi K370.4%estimated ± 2.8 pp, medium confidence
65Claude Opus 4.670.3%estimated ± 2.8 pp, medium confidence
66Muse Spark 1.370.3%estimated ± 2.8 pp, medium confidence
67GPT-570.2%estimated ± 2.8 pp, medium confidence
68GPT-5.270.2%estimated ± 2.8 pp, medium confidence
69Claude Opus 4.570.2%estimated ± 2.8 pp, medium confidence
70Qwen3.6-35B-A3B70.2%estimated ± 2.8 pp, medium confidence
71GPT-5.6 Luna70.1%estimated ± 2.8 pp, medium confidence
72Gemini 3 Flash70.1%estimated ± 2.8 pp, medium confidence
73Nemotron 3 Ultra70.1%estimated ± 2.8 pp, medium confidence
74Qwen3.5 Flash70.1%estimated ± 2.8 pp, medium confidence
75Gemini 3.5 Flash-Lite70.1%estimated ± 2.8 pp, medium confidence
76Qwen 3.6 Max (preview)70.1%estimated ± 2.8 pp, medium confidence
77Kimi K2.670.1%estimated ± 2.8 pp, medium confidence
78Qwen3.5 397B70.1%estimated ± 2.8 pp, medium confidence
79Claude Sonnet 4.570.0%estimated ± 2.8 pp, medium confidence
80Mistral Large 470.0%estimated ± 2.8 pp, medium confidence
81Qwen3.5-122B-A10B70.0%estimated ± 2.8 pp, medium confidence
82Qwen3.5 Plus70.0%estimated ± 2.8 pp, medium confidence
83Qwen3.7 Plus70.0%estimated ± 2.8 pp, medium confidence
84Claude Sonnet 4.670.0%estimated ± 2.8 pp, medium confidence
85GPT-5.170.0%estimated ± 2.8 pp, medium confidence
86GLM-5.270.0%estimated ± 2.8 pp, medium confidence
87Qwen3.7 Flash70.0%estimated ± 2.8 pp, medium confidence
88GPT-4o mini70.0%estimated ± 2.8 pp, medium confidence
89Qwen3.6 Plus70.0%estimated ± 2.8 pp, medium confidence
90MiniMax M370.0%estimated ± 2.8 pp, medium confidence
91GPT-4.1 mini70.0%estimated ± 2.8 pp, low confidence
92GPT-5.4 mini70.0%estimated ± 2.8 pp, low confidence
93GPT-6 Luna70.0%estimated ± 2.8 pp, low confidence
94o3-mini70.0%estimated ± 2.8 pp, low confidence
95Qwen3.6-27B70.0%estimated ± 2.8 pp, low confidence
96Inkling-Small70.0%estimated ± 2.8 pp, low confidence
97GPT-5.4 nano70.0%estimated ± 2.8 pp, low confidence
98o4-mini70.0%estimated ± 2.8 pp, low confidence
99Qwen3 Max70.0%estimated ± 2.8 pp, low confidence
100GPT-OSS 120B70.0%estimated ± 2.8 pp, low confidence
101GLM-5.369.7%measured
102Gemma 4 31B64.8%estimated ± 4.3 pp, low confidence
103Hy4 preview61.0%estimated ± 4.3 pp, low confidence
104Qwen3.8-27B32.2%estimated ± 4.3 pp, low confidence
105Claude Opus 4.5 Thinking27.7%estimated ± 4.3 pp, low confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General