OptiVerse - Medium: leaderboard

OptiVerse (2026) is a benchmark of textbook-sourced optimization problems that models solve by writing and running solver code, a problem counting as solved only when every required value matches the ground truth within 0.1%. This board is accuracy in percent on the medium problems; higher is better.

Metric: Accuracy (%) on the 400 medium-level problems of OptiVerse; the model writes and runs solver code (gurobi, pyomo, cvxpy, ortools and others) and a problem counts as solved only when every required variable and the objective match the ground truth within 0.1% relative error, extracted and checked by a DeepSeek-V3.2-Chat judge; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 20 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3 Flash53.25#180
2Gemini 3 Pro52.75#146
3GPT-5.2Thinking50.75#190
4O347.25#242
5Qwen3-235B-Thinking (OptiVerse checkpoint unspecified)47
6O4 Mini46.75#372
7Claude Sonnet 4.5Thinking45.25#267
8DeepSeek V3.2Thinking44.5#425
9Gemini 2.5 Pro43.75#280
10Gemini 2.5 Flash42.75#466
11Qwen3-30B-Thinking (OptiVerse checkpoint unspecified)40
12Qwen 3 235B Instruct (OptiVerse checkpoint unspecified)39.75
13GPT-OSS-120B39.25#712
14Qwen 3 8BThinking33#1890
15Kimi K231.5#448

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=optiverse-medium · How It Works · Data refreshed daily, snapshot 2026-11-22.