Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Formal Theorem Proving on Combibench
Loading...
48
Solve Rate
Seed-Prover 1.5
-1.7952
11.1324
24.06
36.9876
Dec 19, 2025
Jan 9, 2026
Jan 31, 2026
Feb 22, 2026
Mar 16, 2026
Apr 7, 2026
Apr 29, 2026
Solve Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Solve Rate
Seed-Prover 1.5
Compute Budget=10 H20...
2025.12
48
Seed-Prover 1.0 (medium)
Compute Budget=18 H20...
2025.12
39
Hilbert
Backbone=Gemini 3.1 Pr...
2026.04
2.05
Goedel-Prover-V2-8B
Type=Open-source LLM
2026.04
1.96
Hilbert
Backbone=Gemini 2.5 Pr...
2026.04
1.89
Goedel-Prover-V2-32B
Type=Open-source LLM
2026.04
1.31
DreamProver
Backbone=Gemini 3.1 Pr...
2026.04
0.93
DreamProver
Backbone=Gemini 2.5 Pr...
2026.04
0.75
Gemini 2.5 Pro
Type=Proprietary LLM
2026.04
0.64
Gemini 3.1 Pro
Type=Proprietary LLM
2026.04
0.59
Hilbert
Backbone=GPT-5.3-Codex...
2026.04
0.58
DeepSeek-Prover-V2-7B
Type=Open-source LLM
2026.04
0.42
Claude 4.6 Opus
Type=Proprietary LLM
2026.04
0.41
DreamProver
Backbone=GPT-5.3-Codex...
2026.04
0.32
GPT-5.3-Codex
Type=Proprietary LLM
2026.04
0.12
Feedback
Search any
task
Search any
task