Benchmark Almanac

First Proof Third Batch

ProofCouncil (multiple models) 0.87 tops its leaderboard GPT-6 Astra 0.65 leads the models we track
Not in model cards

The editors run five AI systems once each on ten unpublished research problems with proofs of up to 12 pages. About 30 referees score correctness, exposition and attribution. The score is average correctness, from negative 1 to 1.

Results came out on October 14, 2026. The best harness solved 9 of 10 problems and GPT-6 Astra alone solved 7. A fourth batch is planned for January 2027.

Released
October 2026
Kind
Research mathematics
Level
Research frontier
Scored
Graded by people
Answers
Written proof
Tasks
All public
Strongest model tested
GPT-6 Astra #3 on the capability index
Last model added
September 2026
Scores refreshed
November 12th, 2026
5
Entries on its own leaderboard, 3 models we track

ProofCouncil (multiple models) tops the leaderboard with 0.87

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is GPT-6 Astra, at 0.65.

The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 0.87 ProofCouncil (multiple models)
  2. 0.65 GPT-6 Astra
  3. 0.25 Claude Opus 5.5
  4. 0.23 MerLean (GPT-6, variant unspecified)
  5. -0.86 Kimi K3
  1. 1 ProofCouncil (multiple models) 0.87
  2. 2 GPT-6 Astra 0.65
  3. 3 Claude Opus 5.5 0.25
  4. 4 MerLean (GPT-6, variant unspecified) 0.23
  5. 5 Kimi K3 -0.86

Source: Official First Proof Third Batch leaderboard

#EntryScoreSettingDateSource
1 ProofCouncil (multiple models) 0.87 ProofCouncil Official leaderboard
2 GPT-6 Astra 0.65 Sep 2026 Official leaderboard
3 Claude Opus 5.5 0.25 Siege Sep 2026 Official leaderboard
4 MerLean (GPT-6, variant unspecified) 0.23 MerLean Official leaderboard
5 Kimi K3 -0.86 Humanize Jul 2026 Official leaderboard

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Own leaderboard 1 leader
GPT-6 Astra 0.65

Source: Official First Proof Third Batch leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
-1.00 -0.50 0.00 0.50 1.00 -1.00 -0.50 0.00 0.50 1.00

Source: Official First Proof Third Batch leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

Versions

Replaces First Proof Second Batch

More in Math

All 113 in Math
  • AIME 2026 In use Models solve the 30 integer-answer problems from the 2026 AIME I and II math competitions. Scored by exact answer match. GPT-5.5 100%
  • IMO-AnswerBench In use 400 olympiad problems with short verifiable answers, robustified against memorization. Scored by answer match. GLM-5.2 91%
  • HMMT 2026 In use Models solve the 33 final-answer problems from the February 2026 Harvard-MIT Mathematics Tournament. Scored by exact answer match. GPT-5.5 98.5%
  • PolyMath In use Models solve 9,000 parallel math problems in 18 languages across four difficulty levels, from school math to frontier math. Final answers are checked. Qwen3.7-Max 86.5%
  • MathArena (Apex, IMO, USAMO) In use Evaluates models on newly released math competitions after release. Final-answer or human/LLM-judged proof scores. Claude Opus 4.8 81.3%
  • MathArena Apex Shortlist In use Models solve 47 hard final-answer problems that MathArena selected from 2025 competitions, where the best models scored about 50% at selection. Scored by accuracy over repeated attempts. GPT-5.5 98.4%