Benchmark Almanac

TxBench-Antibody-Discovery

Claude Opus 5.5 67.7% leads since September 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
August 2026
Kind
Genomics and life sciences
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
Public, plus a private test set
Strongest model tested
Gemini 4 Carbon #2 on the capability index
Last model added
November 2026
Scores refreshed
November 13th, 2026
30
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 November 13th, 2026. Where a model ran with several settings or harnesses, we keep its best.

30 models. Scroll sideways for the rest.

  1. 67.7 Claude Opus 5.5
  2. 65.0 Gemini 4 Carbon
  3. 63.7 Gemini 4 Argon
  4. 63.3 Claude Fable 5.1
  5. 57.7 Gemini 4 Hydrogen
  6. 57.3 Claude Sonnet 5.5
  7. 56.0 GPT-6 Astra
  8. 53.0 Grok 4.6
  9. 52.0 Grok 4.7
  10. 50.7 Gemini 3.7 Flash
  11. 49.7 GPT-6.1 Sol
  12. 49.0 Gemini 3.8 Flash
  13. 48.7 Claude Opus 5
  14. 43.3 GPT-6 Sol
  15. 40.7 GPT-6.1 Terra
  16. 39.7 DeepSeek V4.1 Flash
  17. 37.3 GPT-5.6 Sol
  18. 36.0 Claude Opus 4.7
  19. 35.3 Claude Opus 4.8
  20. 35.3 Kimi K3
  21. 34.0 GPT-6 Luna
  22. 34.0 Beam 1.0
  23. 33.0 GPT-5.5
  24. 32.0 Claude Sonnet 5
  25. 28.3 Mistral Large 4
  26. 26.3 Claude Sonnet 4.6
  27. 26.3 GLM-5.3
  28. 21.0 Nemotron 3 Ultra
  29. 19.3 GPT-5.6 Luna
  30. 16.0 Nemotron 3.5 Lightning 30B-A3B
  1. 1 Claude Opus 5.5 LSVP 67.7%
  2. 2 Gemini 4 Carbon 65%
  3. 3 Gemini 4 Argon 63.7%
  4. 4 Claude Fable 5.1 LSVP 63.3%
  5. 5 Gemini 4 Hydrogen 57.7%
  6. 6 Claude Sonnet 5.5 57.3%
  7. 7 GPT-6 Astra 56%
  8. 8 Grok 4.6 53%
  9. 9 Grok 4.7 52%
  10. 10 Gemini 3.7 Flash 50.7%
  11. 11 GPT-6.1 Sol 49.7%
  12. 12 Gemini 3.8 Flash 49%
  13. 13 Claude Opus 5 48.7%
  14. 14 GPT-6 Sol 43.3%
  15. 15 GPT-6.1 Terra 40.7%
  16. 16 DeepSeek V4.1 Flash 39.7%
  17. 17 GPT-5.6 Sol 37.3%
  18. 18 Claude Opus 4.7 36%
  19. 19 Claude Opus 4.8 35.3%
  20. 20 Kimi K3 35.3%
  21. 21 GPT-6 Luna 34%
  22. 22 Beam 1.0 34%
  23. 23 GPT-5.5 33%
  24. 24 Claude Sonnet 5 32%
  25. 25 Mistral Large 4 28.3%
  26. 26 Claude Sonnet 4.6 26.3%
  27. 27 GLM-5.3 26.3%
  28. 28 Nemotron 3 Ultra 21%
  29. 29 GPT-5.6 Luna 19.3%
  30. 30 Nemotron 3.5 Lightning 30B-A3B 16%

Source: Official TxBench-Antibody-Discovery leaderboard

#ModelScoreSettingDateSource
1 Claude Opus 5.5 67.7% LSVP, Claude Code Sep 2026 Official leaderboard
2 Gemini 4 Carbon 65% Pi Oct 2026 Official leaderboard
3 Gemini 4 Argon 63.7% Pi Sep 2026 Official leaderboard
4 Claude Fable 5.1 63.3% LSVP, Pi Sep 2026 Official leaderboard
5 Gemini 4 Hydrogen 57.7% Pi Nov 2026 Official leaderboard
6 Claude Sonnet 5.5 57.3% Claude Code Sep 2026 Official leaderboard
7 GPT-6 Astra 56% Pi Sep 2026 Official leaderboard
8 Grok 4.6 53% Pi Aug 2026 Official leaderboard
9 Grok 4.7 52% Pi Sep 2026 Official leaderboard
10 Gemini 3.7 Flash 50.7% Pi Aug 2026 Official leaderboard
11 GPT-6.1 Sol 49.7% Pi Sep 2026 Official leaderboard
12 Gemini 3.8 Flash 49% Pi Sep 2026 Official leaderboard
13 Claude Opus 5 48.7% Pi Jul 2026 Official leaderboard
14 GPT-6 Sol 43.3% Pi Sep 2026 Official leaderboard
15 GPT-6.1 Terra 40.7% Pi Oct 2026 Official leaderboard
16 DeepSeek V4.1 Flash 39.7% Pi Sep 2026 Official leaderboard
17 GPT-5.6 Sol 37.3% Pi Jul 2026 Official leaderboard
18 Claude Opus 4.7 36% Claude Code Apr 2026 Official leaderboard
19 Claude Opus 4.8 35.3% Claude Code May 2026 Official leaderboard
20 Kimi K3 35.3% Pi Jul 2026 Official leaderboard
21 GPT-6 Luna 34% Pi Sep 2026 Official leaderboard
22 Beam 1.0 34% Pi Oct 2026 Official leaderboard
23 GPT-5.5 33% Pi Apr 2026 Official leaderboard
24 Claude Sonnet 5 32% Claude Code Jun 2026 Official leaderboard
25 Mistral Large 4 28.3% Pi Oct 2026 Official leaderboard
26 Claude Sonnet 4.6 26.3% Pi Feb 2026 Official leaderboard
27 GLM-5.3 26.3% Pi Aug 2026 Official leaderboard
28 Nemotron 3 Ultra 21% Pi Jun 2026 Official leaderboard
29 GPT-5.6 Luna 19.3% Pi Jul 2026 Official leaderboard
30 Nemotron 3.5 Lightning 30B-A3B 16% Pi Aug 2026 Official leaderboard

4 models have held number one on the official leaderboard since its release in August 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%

Source: Official TxBench-Antibody-Discovery leaderboard

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

Source: Official TxBench-Antibody-Discovery leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Biology and chemistry

All 108 in Biology and chemistry
  • 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-5.6 Sol 31.5%
  • 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%
  • 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-Rosalind-6 61.3%
  • MedChemBench (OpenAI) In use Models answer chemistry questions from an OpenAI internal set, grouped with OpenAI's life science research evaluations. OpenAI has not published the task format. GPT-Rosalind-6 49.6%
  • ProtocolQA Open-Ended In use Models correct errors in published lab protocols using open-ended questions adapted from LAB-Bench ProtocolQA. Scored as pass@1. GPT-6 Cyber 47.2%
  • MolecularIQ In use Models answer questions about molecular graphs, such as counting, indexing and constrained generation, and a symbolic checker verifies each answer. gpt-oss-120b 47.5%