Benchmark Almanac

FoamBench V2

Gemini 3.1 Pro 86.5% leads since October 2026
Not in model cards
Version 2

Agents build and run 110 Basic and 16 Advanced OpenFOAM cases. Revised instructions specify geometry, meshes, and boundary conditions for the same named cases.

The authors score six model backends and two standalone coding agents. They report execution and physical success separately for each split.

Released
October 2026
Kind
Apps and MCP servers
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Code, Actions
Interaction
Agentic
Environment
Terminal, OpenFOAM
Run by
Anyone
Based on
FoamBench
In views
Agents
Strongest model tested
Claude Opus 5.5 #5 on the capability index
Last model added
September 2026
Scores refreshed
November 20th, 2026
7
Models scored on its own leaderboard

Gemini 3.1 Pro leads with 86.5% among models released by September 2026

Best score per model from the benchmark's own leaderboard, read on November 20th, 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 Gemini 3.1 Pro Low reasoning effort 86.5%
  2. 2 Gemini 3.8 Flash Low reasoning effort 79.7%
  3. 3 GPT-5.6 Sol Low reasoning effort 75.3%
  4. 4 GLM-5.2 Thinking off 58.3%
  5. 5 Claude Opus 5.5 Low reasoning effort 54.5%
  6. 6 Qwen 3.8 27B Thinking off 44.6%
  7. 7 Llama 4 Maverick 24%

Source: Official FoamBench V2 leaderboard

#ModelScoreSettingDateSource
1 Gemini 3.1 Pro 86.5% Low reasoning effort, Foam-Agent 2.0 Feb 2026 Official leaderboard
2 Gemini 3.8 Flash 79.7% Low reasoning effort, Foam-Agent 2.0 Sep 2026 Official leaderboard
3 GPT-5.6 Sol 75.3% Low reasoning effort, Foam-Agent 1.0 Jul 2026 Official leaderboard
4 GLM-5.2 58.3% Thinking off, Foam-Agent 1.0 Jun 2026 Official leaderboard
5 Claude Opus 5.5 54.5% Low reasoning effort, Claude Code Sep 2026 Official leaderboard
6 Qwen 3.8 27B 44.6% Thinking off, Foam-Agent 2.0 Aug 2026 Official leaderboard
7 Llama 4 Maverick 24% Foam-Agent 1.0 Apr 2025 Official leaderboard

Rankings by part

We read 2 parts from the benchmark’s own leaderboard: Basic and Advanced. Each ranks models on one part of the test. Gemini 3.1 Pro leads both. Arrows show how far a model moved from its overall rank among the models we track.

Basic

7 models we track have a score here.

  1. 1 Gemini 3.1 Pro 81.8%
  2. 2 GLM-5.2 ↑2 79.1%
  3. 3 Gemini 3.8 Flash ↓1 78.2%
  4. 4 GPT-5.6 Sol ↓1 75.5%
  5. 5 Qwen 3.8 27B ↑1 62.7%
  6. 6 Claude Opus 5.5 ↓1 52.7%
  7. 7 Llama 4 Maverick 41.8%

Number one over time

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

Google DeepMind
Own leaderboard Passed 85% in Oct 2026, by Gemini 3.1 Pro

Source: Official FoamBench V2 leaderboard

Gemini 3.1 Pro on the official leaderboard passed 85% in February 2026

Models released before it came out already scored that high. 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 Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Gemini 3.1 Pro Llama 4 Maverick 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Gemini 3.1 Pro Llama 4 Maverick

Source: Official FoamBench V2 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.

Versions

We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.

OpenAI Google DeepMind
FoamBench V2 Passed 85% in Oct 2026, by Gemini 3.1 Pro
FoamBench Not tested since Sep 2025
o3-mini 22.5%
FoamBench V2 Passed 85% in Oct 2026, by Gemini 3.1 Pro

Source: Official leaderboards

More in Tool use

All 265 in Tool use
  • AutomationBench In use Agent completes end-to-end business workflows inside a simulated company exposing dozens of REST APIs. Graded on resulting state. Gemini 4 Neon 56.3%
  • MCP Atlas In use Multi-step workflows over real MCP servers. Scores tool-use competency on realistic tasks. Gemini 4 Argon 90.7%
  • BFCL V4 In use Models call functions and tools, including multi-turn agentic tasks with web search and memory. Checked against expected calls and outcomes. Qwen3.7-Max 75%
  • τ³-bench Banking In use Agents act as banking customer-service reps over many turns with a simulated user, following policy. Pass^1 on the final database state. Qwen 3.8 Max 55.1%
  • Toolathlon-Verified In use 108 long-horizon tasks across 32 apps through about 600 MCP tools. Execution-based checkers inspect final environment state. GLM-5.3-Flash 78.4%
  • DeepSearchQA In use Questions that need multi-step search and synthesis. Final answers graded against references. Kimi K3 95%

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 Claude Opus 5.5.
  • Note Frontier models 1 of the top 10 has a score, which is fine for a retired test. The strongest tested is Claude Opus 5.5 (#5).
  • Note Headroom Saturated. The best is 86.5%, past 85% since February 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on nithinsomu.github.io: 14 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