Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Text-to-code generation on DTVBENCH
Loading...
10.6
Manim Score
GPT-4o
4.6928
6.2264
7.76
9.2936
Oct 27, 2025
Manim Score
Wolfram Score
Updated 4d ago
Evaluation Results
Method
Method
Links
Manim Score
Wolfram Score
GPT-4o
Model Type=Proprietary
2025.10
10.6
4.92
JANUSCODER-8B
Model Type=Open-Source...
2025.10
9.7
6.07
Qwen2.5-Coder-32B-Ins
Model Type=Open-Source...
2025.10
9.61
4.98
Qwen2.5-Coder-7B-Ins
Model Type=Open-Source...
2025.10
8.56
4.04
JANUSCODER-14B
Model Type=Open-Source...
2025.10
8.41
5.97
Qwen3-14B
Model Type=Open-Source...
2025.10
6.63
5.08
Qwen3-8B
Model Type=Open-Source...
2025.10
6.2
5.18
LLaMA3-8B-Instruct
Model Type=Open-Source...
2025.10
4.92
3.15
Feedback
Search any
task
Search any
task