OptiVerse - Optimal Control: 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 optimal control problems; higher is better.

Metric: Accuracy (%) on the 78 optimal control 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
1GPT-5.2Thinking50#185
2Gemini 2.5 Flash48.72#463
3Gemini 3 Flash48.72#175
4Qwen3-235B-Thinking (OptiVerse checkpoint unspecified)43.59
5Gemini 3 Pro42.31#145
6O4 Mini42.31#368
7Qwen 3 235B Instruct (OptiVerse checkpoint unspecified)42.31
8O339.74#237
9Gemini 2.5 Pro38.46#276
10Claude Sonnet 4.5Thinking34.62#261
11GPT-OSS-120B32.05#704
12Qwen3-30B-Thinking (OptiVerse checkpoint unspecified)30.77
13DeepSeek V3.2Thinking28.21#419
14Qwen 3 8BThinking25.64#1873
15Kimi K224.36#443

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=optiverse-optimal-control · How It Works · Data refreshed daily, snapshot 2026-11-21.