Benchmark Almanac

TxBench-Antibody-Discovery

Claude Opus 5.5 67.7% leads since Sep 2026 Not in model cards

Agents make antibody discovery and development decisions from real experimental data. A grader checks each decision.

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 Aug 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/txbench-antibody-discovery 7 stars Leaderboard benchmarks.bio
28
Models scored on its own leaderboard

Claude Opus 5.5 leads with 67.7%

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.

28 models. Scroll sideways for the rest.

  1. 67.7 Claude Opus 5.5
  2. 63.7 Gemini 4 Argon
  3. 63.3 Claude Fable 5.1
  4. 57.3 Claude Sonnet 5.5
  5. 56.0 GPT-6 Astra
  6. 53.0 Grok 4.6
  7. 52.0 Grok 4.7
  8. 50.7 Gemini 3.7 Flash
  9. 49.7 GPT-6.1 Sol
  10. 49.0 Gemini 3.8 Flash
  11. 48.7 Claude Opus 5
  12. 43.3 GPT-6 Sol
  13. 40.7 GPT-6.1 Terra
  14. 39.7 DeepSeek V4.1 Flash
  15. 37.3 GPT-5.6 Sol
  16. 36.0 Claude Opus 4.7
  17. 35.3 Claude Opus 4.8
  18. 35.3 Kimi K3
  19. 34.0 GPT-6 Luna
  20. 34.0 Beam 1.0
  21. 33.0 GPT-5.5
  22. 32.0 Claude Sonnet 5
  23. 28.3 Mistral Large 4
  24. 26.3 Claude Sonnet 4.6
  25. 26.3 GLM-5.3
  26. 21.0 Nemotron 3 Ultra
  27. 19.3 GPT-5.6 Luna
  28. 16.0 Nemotron 3.5 Lightning 30B-A3B
  1. 1 Claude Opus 5.5 LSVP 67.7%
  2. 2 Gemini 4 Argon 63.7%
  3. 3 Claude Fable 5.1 LSVP 63.3%
  4. 4 Claude Sonnet 5.5 57.3%
  5. 5 GPT-6 Astra 56%
  6. 6 Grok 4.6 53%
  7. 7 Grok 4.7 52%
  8. 8 Gemini 3.7 Flash 50.7%
  9. 9 GPT-6.1 Sol 49.7%
  10. 10 Gemini 3.8 Flash 49%
  11. 11 Claude Opus 5 48.7%
  12. 12 GPT-6 Sol 43.3%
  13. 13 GPT-6.1 Terra 40.7%
  14. 14 DeepSeek V4.1 Flash 39.7%
  15. 15 GPT-5.6 Sol 37.3%
  16. 16 Claude Opus 4.7 36%
  17. 17 Claude Opus 4.8 35.3%
  18. 18 Kimi K3 35.3%
  19. 19 GPT-6 Luna 34%
  20. 20 Beam 1.0 34%
  21. 21 GPT-5.5 33%
  22. 22 Claude Sonnet 5 32%
  23. 23 Mistral Large 4 28.3%
  24. 24 Claude Sonnet 4.6 26.3%
  25. 25 GLM-5.3 26.3%
  26. 26 Nemotron 3 Ultra 21%
  27. 27 GPT-5.6 Luna 19.3%
  28. 28 Nemotron 3.5 Lightning 30B-A3B 16%
#ModelScoreSettingDateSource
1 Claude Opus 5.5 67.7% LSVP, Claude Code Sep 2026 Official leaderboard
2 Gemini 4 Argon 63.7% Pi Sep 2026 Official leaderboard
3 Claude Fable 5.1 63.3% LSVP, Pi Sep 2026 Official leaderboard
4 Claude Sonnet 5.5 57.3% Claude Code Sep 2026 Official leaderboard
5 GPT-6 Astra 56% Pi Sep 2026 Official leaderboard
6 Grok 4.6 53% Pi Aug 2026 Official leaderboard
7 Grok 4.7 52% Pi Sep 2026 Official leaderboard
8 Gemini 3.7 Flash 50.7% Pi Aug 2026 Official leaderboard
9 GPT-6.1 Sol 49.7% Pi Sep 2026 Official leaderboard
10 Gemini 3.8 Flash 49% Pi Sep 2026 Official leaderboard
11 Claude Opus 5 48.7% Pi Jul 2026 Official leaderboard
12 GPT-6 Sol 43.3% Pi Sep 2026 Official leaderboard
13 GPT-6.1 Terra 40.7% Pi Oct 2026 Official leaderboard
14 DeepSeek V4.1 Flash 39.7% Pi Sep 2026 Official leaderboard
15 GPT-5.6 Sol 37.3% Pi Jul 2026 Official leaderboard
16 Claude Opus 4.7 36% Claude Code Apr 2026 Official leaderboard
17 Claude Opus 4.8 35.3% Claude Code May 2026 Official leaderboard
18 Kimi K3 35.3% Pi Jul 2026 Official leaderboard
19 GPT-6 Luna 34% Pi Sep 2026 Official leaderboard
20 Beam 1.0 34% Pi Oct 2026 Official leaderboard
21 GPT-5.5 33% Pi Apr 2026 Official leaderboard
22 Claude Sonnet 5 32% Claude Code Jun 2026 Official leaderboard
23 Mistral Large 4 28.3% Pi Oct 2026 Official leaderboard
24 Claude Sonnet 4.6 26.3% Pi Feb 2026 Official leaderboard
25 GLM-5.3 26.3% Pi Aug 2026 Official leaderboard
26 Nemotron 3 Ultra 21% Pi Jun 2026 Official leaderboard
27 GPT-5.6 Luna 19.3% Pi Jul 2026 Official leaderboard
28 Nemotron 3.5 Lightning 30B-A3B 16% Pi Aug 2026 Official leaderboard

4 models have held number one since its release in Aug 2026

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

Anthropic xAI
Own leaderboard 4 leaders
Claude Opus 5.5 67.7%

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 Fable 5.1 Grok 4.6 Claude Opus 5 GPT-5.6 Sol Claude Opus 4.7 Claude Sonnet 4.6 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Opus 5.5 Grok 4.6

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%