BlogLLM Benchmark Scores as Published, Side by Side
MiniF2F: Results as Published
Every result the catalog holds for MiniF2F: 3 models from 3 developers, read from 1 page. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Science and MathMaintainer Kunhao Zheng, Jesse Michael Han, and Stanislas Polu (repository published under openai)Introduced 2021Data version 2026103001, as of 2026-10-30
MiniF2F is a formal mathematics benchmark for neural theorem proving. It has 488 problem statements drawn from AIME, AMC, the International Mathematical Olympiad, and high-school and undergraduate courses, written for Metamath, Lean, and partially for Isabelle and HOL Light. The model must produce a proof that the proof assistant accepts. The paper reports the pass rate on a held-out test split, including Pass@1 and Pass@8.
MiniF2F (pass@1)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| gpt-oss-20bOpenAI, released 2025-08-05 | 12.1% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| Qwen3-30B-A3B-Thinking-2507Alibaba, released 2025-07-30 | 5.7% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Effort: thinking |
| Nemotron 3 Nano 30B-A3BNVIDIA, released 2025-12-15 | 50% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
MiniF2F (pass@32)
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| gpt-oss-20bOpenAI, released 2025-08-05 | 43% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
| Qwen3-30B-A3B-Thinking-2507Alibaba, released 2025-07-30 | 16.8% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | Effort: thinking |
| Nemotron 3 Nano 30B-A3BNVIDIA, released 2025-12-15 | 79.9% | NVIDIA, NVIDIA Nemotron 3 Nano 30B-A3B model cardchecked 2026-10-15 | - |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.