Benchmark Almanac

MetagenomicsBench

Claude Opus 5.5 62.7% leads since Sep 2026 Not in model cards

Agents analyze microbial communities from shotgun, amplicon, long-read and spatial metagenomics data. A grader checks the biological result.

Part of LatchBio's BioBench suite on benchmarks.bio. The V1 set has 100 evaluations. Release month is the GitHub repo creation month.

Made by LatchBio Released Sep 2026 Genomics and life sciences Public tasks plus a private test set Last model added Oct 2026 Scores refreshed 2 Nov 2026
Level
Expert Specialists with years of training, often over hours of work
Scored
Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
Answers
Agent task Many steps in a terminal, browser, desktop, app, game or robot.
latchbio/metagenomicsbench 1 stars Leaderboard benchmarks.bio
26
Models scored on its own leaderboard

Claude Opus 5.5 leads with 62.7%

Best score per model from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best.

26 models. Scroll sideways for the rest.

  1. 62.7 Claude Opus 5.5
  2. 61.7 Claude Fable 5.1
  3. 60.0 Claude Sonnet 5.5
  4. 54.0 Claude Opus 5
  5. 50.0 Claude Opus 4.8
  6. 49.7 GPT-6 Astra
  7. 48.7 Grok 4.7
  8. 45.7 GPT-6.1 Sol
  9. 42.3 Claude Sonnet 5
  10. 42.3 Gemini 4 Argon
  11. 42.0 Grok 4.6
  12. 42.0 GPT-6 Sol
  13. 38.7 GPT-6.1 Terra
  14. 38.0 DeepSeek V4.1 Flash
  15. 35.3 Claude Opus 4.7
  16. 31.7 Gemini 3.7 Flash
  17. 31.3 GPT-5.6 Sol
  18. 31.0 GPT-6 Luna
  19. 28.3 Claude Sonnet 4.6
  20. 27.3 Kimi K3
  21. 27.3 GLM-5.3
  22. 26.7 Gemini 3.8 Flash
  23. 26.3 Mistral Large 4
  24. 24.3 GPT-5.6 Luna
  25. 16.3 Nemotron 3 Ultra
  26. 7.67 Nemotron 3.5 Lightning 30B-A3B
  1. 1 Claude Opus 5.5 LSVP 62.7%
  2. 2 Claude Fable 5.1 LSVP 61.7%
  3. 3 Claude Sonnet 5.5 60%
  4. 4 Claude Opus 5 54%
  5. 5 Claude Opus 4.8 50%
  6. 6 GPT-6 Astra 49.7%
  7. 7 Grok 4.7 48.7%
  8. 8 GPT-6.1 Sol 45.7%
  9. 9 Claude Sonnet 5 42.3%
  10. 10 Gemini 4 Argon 42.3%
  11. 11 Grok 4.6 42%
  12. 12 GPT-6 Sol 42%
  13. 13 GPT-6.1 Terra 38.7%
  14. 14 DeepSeek V4.1 Flash 38%
  15. 15 Claude Opus 4.7 35.3%
  16. 16 Gemini 3.7 Flash 31.7%
  17. 17 GPT-5.6 Sol 31.3%
  18. 18 GPT-6 Luna 31%
  19. 19 Claude Sonnet 4.6 28.3%
  20. 20 Kimi K3 27.3%
  21. 21 GLM-5.3 27.3%
  22. 22 Gemini 3.8 Flash 26.7%
  23. 23 Mistral Large 4 26.3%
  24. 24 GPT-5.6 Luna 24.3%
  25. 25 Nemotron 3 Ultra 16.3%
  26. 26 Nemotron 3.5 Lightning 30B-A3B 7.67%
#ModelScoreSettingDateSource
1 Claude Opus 5.5 62.7% LSVP, Claude Code Sep 2026 Official leaderboard
2 Claude Fable 5.1 61.7% LSVP, Pi Sep 2026 Official leaderboard
3 Claude Sonnet 5.5 60% Claude Code Sep 2026 Official leaderboard
4 Claude Opus 5 54% Pi Jul 2026 Official leaderboard
5 Claude Opus 4.8 50% Claude Code May 2026 Official leaderboard
6 GPT-6 Astra 49.7% Codex Sep 2026 Official leaderboard
7 Grok 4.7 48.7% Pi Sep 2026 Official leaderboard
8 GPT-6.1 Sol 45.7% Pi Sep 2026 Official leaderboard
9 Claude Sonnet 5 42.3% Pi Jun 2026 Official leaderboard
10 Gemini 4 Argon 42.3% Pi Sep 2026 Official leaderboard
11 Grok 4.6 42% Grok Build Aug 2026 Official leaderboard
12 GPT-6 Sol 42% Pi Sep 2026 Official leaderboard
13 GPT-6.1 Terra 38.7% Pi Oct 2026 Official leaderboard
14 DeepSeek V4.1 Flash 38% Pi Sep 2026 Official leaderboard
15 Claude Opus 4.7 35.3% Pi Apr 2026 Official leaderboard
16 Gemini 3.7 Flash 31.7% Pi Aug 2026 Official leaderboard
17 GPT-5.6 Sol 31.3% Pi Jul 2026 Official leaderboard
18 GPT-6 Luna 31% Codex Sep 2026 Official leaderboard
19 Claude Sonnet 4.6 28.3% Claude Code Feb 2026 Official leaderboard
20 Kimi K3 27.3% Pi Jul 2026 Official leaderboard
21 GLM-5.3 27.3% Pi Aug 2026 Official leaderboard
22 Gemini 3.8 Flash 26.7% Pi Sep 2026 Official leaderboard
23 Mistral Large 4 26.3% Pi Oct 2026 Official leaderboard
24 GPT-5.6 Luna 24.3% Codex Jul 2026 Official leaderboard
25 Nemotron 3 Ultra 16.3% Pi Jun 2026 Official leaderboard
26 Nemotron 3.5 Lightning 30B-A3B 7.67% Pi Aug 2026 Official leaderboard

2 models have held number one since its release in Sep 2026

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

Anthropic
Own leaderboard 2 leaders
Claude Opus 5.5 62.7%

Scores over time

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

On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Opus 5.5 Claude Opus 5 Claude Opus 4.7 Claude Sonnet 4.6 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Opus 5.5 Claude Opus 5

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Biology and chemistry

All 96 in Biology and chemistry
  • SUPERChem In use Models solve 500 expert-written chemistry reasoning problems, offered in multimodal and text-only forms. Expert solution paths also allow a reasoning path score. Tencent Hy4 preview 66.4%
  • GeneBench Pro In use Agents analyze messy simulated genomics and biology datasets to estimate a decision-relevant quantity across 129 problems. Scored automatically against known answers. GPT-6 Astra 37.1%
  • LifeSciBench In use Models complete 750 expert-written life science research tasks, often with attached figures, tables, or sequence files. Graded against expert rubrics as score and pass rate. GPT-6 Astra 60.3%
  • MolecularIQ In use Models answer questions about molecular graphs, such as counting, indexing and constrained generation, and a symbolic checker verifies each answer. Intern-S2-397B 62.4%
  • TOMG-Bench In use Models generate molecules from natural language requests in three tasks: molecule editing, molecule optimization and customized generation. Each output is checked for validity and for meeting the requested properties. Intern-S2-397B 66.8%
  • BABE Fading Models reason from experimental results in peer-reviewed biology papers to draw conclusions, with causal and cross-scale inference. Seed2.0 Lite (0428) 57.9%