Benchmark Almanac

SimulCost

GPT-5 80.8% leads since February 2026
Not in model cards

Agents tune physics simulations to meet an accuracy target. The authors measure success and computational cost for an initial guess and for repeated attempts with feedback.

The authors publish results for five models on 11 simulators with analytical costs. They report the elapsed time for a separate plasma simulation.

Released
February 2026
Kind
Research science
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Efficiency
Language
English
In and out
Text → Actions, Text
Interaction
Agentic
Environment
A simulator
Items from
Generated
Refresh
One fixed set
Run by
Anyone
In views
Agents, Robotics
Strongest model tested
GPT-5 #71 on the capability index
Last model added
August 2025
Scores refreshed
November 20th, 2026
5
Models scored in its paper

GPT-5 leads with 80.8% among models released by August 2025

Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

The newest model with a score here came out in August 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 GPT-5 Multi-round 80.8%
  2. 2 gpt-oss-120b Multi-round 77.1%
  3. 3 Llama 3-70B Multi-round 70.6%
  4. 4 Qwen3-32B Multi-round 67.7%
  5. 5 Claude 3.7 Sonnet Multi-round 66.1%

Source: The SimulCost paper

#ModelScoreSettingDateSource
1 GPT-5 80.8% Multi-round Aug 2025 Official leaderboard
2 gpt-oss-120b 77.1% Multi-round Aug 2025 Official leaderboard
3 Llama 3-70B 70.6% Multi-round Apr 2024 Official leaderboard
4 Qwen3-32B 67.7% Multi-round Apr 2025 Official leaderboard
5 Claude 3.7 Sonnet 66.1% Multi-round Feb 2025 Official leaderboard

Rankings by part

We read 7 parts from the benchmark’s paper, each ranking models on one part of the test, such as Low accuracy success, Medium accuracy success and High accuracy success. GPT-5 leads 4 of them, and gpt-oss-120b leads the other 3. Arrows show how far a model moved from its overall rank among the models we track.

Low accuracy success

5 models we track have a score here.

  1. 1 GPT-5 85.5%
  2. 2 gpt-oss-120b 74.2%
  3. 3 Qwen3-32B ↑1 71%
  4. 4 Llama 3-70B ↓1 69.2%
  5. 5 Claude 3.7 Sonnet 65.6%

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-5
Single-round Multi-round
80.8%
gpt-oss-120b
Single-round Multi-round
77.1%
Llama 3-70B
Single-round Multi-round
70.6%
Qwen3-32B
Single-round Multi-round
67.7%
Claude 3.7 Sonnet
Single-round Multi-round
66.1%

Source: The SimulCost paper

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-5 Multi-round 80.8%
gpt-oss-120b Multi-round 77.1%
Llama 3-70B Multi-round 70.6%
Qwen3-32B Multi-round 67.7%
Claude 3.7 Sonnet Multi-round 66.1%
GPT-5 Single-round 61.5%
Claude 3.7 Sonnet Single-round 52.7%
gpt-oss-120b Single-round 48.5%
Qwen3-32B Single-round 46.3%
Llama 3-70B Single-round 44.8%

Number one over time

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

OpenAI
Benchmark paper Not tested since Feb 2026
GPT-5 80.8%

Source: The SimulCost paper

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
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) GPT-5 Llama 3-70B 0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) GPT-5 Llama 3-70B

Source: Official SimulCost 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 Science and reasoning

All 203 in Science and reasoning
  • GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics. Gemini 4 Carbon 97%
  • Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
  • BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
  • FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics. Intern-S2-397B 87%
  • HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels. Gemini 3 Pro 78.3%
  • LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 15 months ago, in August 2025, for GPT-5. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (5 entries).
  • Pass Models scored 5 models we track.
  • Warning Newest score The newest came 15 months ago, in August 2025, for GPT-5.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5 (#71).
  • Pass Headroom The best is 80.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 10 rows for models we track at the last read.
  • Pass Last read November 20th, 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