Benchmark Almanac

Verify Gym

DeepSeek V4.1 Flash 0.10 leads since October 2026
Not in model cards

Models give tighter rational bounds for roots on 120 held-out tasks. Exact integer arithmetic verifies both bounds and determines how much precision the answer adds.

The published scorecard reports reference policies on 120 held-out tasks. It gives no model score.

Made by
carlos-toledo
Released
October 2026
Kind
Word problems
Level
University
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Subject
Mathematics
Language
English
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
DeepSeek V4.1 Flash #34 on the capability index
Last model added
September 2026
Scores refreshed
November 21st, 2026
7
Models scored on its own leaderboard

DeepSeek V4.1 Flash leads with 0.10 among models released by September 2026

Best score per model from the benchmark's own leaderboard, read on November 21st, 2026. Where a model ran with several settings or harnesses, we keep its best.

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. 1 DeepSeek V4.1 Flash 120 held-out tasks, 8 samples per task, no tools 0.10
  2. 2 gpt-oss-120b 120 held-out tasks, 8 samples per task, no tools 0.09
  3. 3 Gemma 4 26B A4B 120 held-out tasks, 8 samples per task, no tools 0.08
  4. 4 Qwen3.5-35B-A3B 120 held-out tasks, 8 samples per task, no tools 0.08
  5. 5 Qwen3.5-9B 120 held-out tasks, 8 samples per task, no tools 0.08
  6. 6 gpt-oss-20b 120 held-out tasks, 8 samples per task, no tools 0.05
  7. 7 DeepSeek-V4-Pro 120 held-out tasks, 8 samples per task, no tools 0.05

Source: Official Verify Gym leaderboard

#ModelScoreSettingDateSource
1 DeepSeek V4.1 Flash 0.10 120 held-out tasks, 8 samples per task, no tools Sep 2026 Official leaderboard
2 gpt-oss-120b 0.09 120 held-out tasks, 8 samples per task, no tools Aug 2025 Official leaderboard
3 Gemma 4 26B A4B 0.08 120 held-out tasks, 8 samples per task, no tools Apr 2026 Official leaderboard
4 Qwen3.5-35B-A3B 0.08 120 held-out tasks, 8 samples per task, no tools Feb 2026 Official leaderboard
5 Qwen3.5-9B 0.08 120 held-out tasks, 8 samples per task, no tools Feb 2026 Official leaderboard
6 gpt-oss-20b 0.05 120 held-out tasks, 8 samples per task, no tools Aug 2025 Official leaderboard
7 DeepSeek-V4-Pro 0.05 120 held-out tasks, 8 samples per task, no tools Apr 2026 Official leaderboard

Number one over time

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

DeepSeek
Own leaderboard Not tested since Oct 2026

Source: Official Verify Gym leaderboard

Scores over time

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

Plot with other benchmarks
On the benchmark’s own leaderboard
0.06 0.08 0.10 2026 0.06 0.08 0.10 2026

Source: Official Verify Gym leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Math

All 140 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%

Data audit

All 9 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (7 entries).
  • Pass Models scored 7 models we track.
  • Pass Newest score The newest is from September 2026, for DeepSeek V4.1 Flash.
  • Note Frontier models None of the top 10 has a score. The strongest tested is DeepSeek V4.1 Flash (#34).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 7 rows for models we track at the last read.
  • Pass Last read November 21st, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check