Benchmark Almanac

scBench-Long (original study)

Claude Opus 4.8 25.4% leads since June 2026
Not in model cards

Agents complete 21 tasks in long single-cell data analysis workflows. The authors compare their answers with results from the original analyses.

LatchBio distinguishes the original 21 tasks from the current 22. Its current board lists 45 model and harness configurations.

Made by
LatchBio
Released
June 2026
Kind
Genomics and life sciences
Level
Expert
Scored
Checked automatically
Answers
Agent task
Tasks
Public, plus a private test set
Measures
Capability
Language
English
In and out
Text, Code → Text, Actions
Interaction
Agentic
Environment
Code sandbox
Items from
Real work and use
Run by
A host
Strongest model tested
GPT-5.5 #21 on the capability index
Last model added
May 2026
Scores refreshed
November 21st, 2026
12
Entries in its paper, 10 models we track

Claude Opus 4.8 leads with 25.4% among models released by May 2026 Claude Opus 4.8 leads with 25.4% among models released by May 2026

Best score per entry from the benchmark's paper, read on November 21st, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in May 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 Claude Opus 4.8 25.4%
  2. 2 Gemini 3.5 Flash 22.2%
  3. 3 Gemini 3.1 Pro 17.5%
  4. 4 GPT-5.5 15.9%
  5. 5 Claude Opus 4.7 11.1%
  6. 6 GLM-5P2 9.52%
  7. 7 Claude Opus 4.6 9.52%
  8. 8 GPT-5.4 9.52%
  9. 9 Claude Sonnet 4.6 7.94%
  10. 10 Grok 4.20 6.35%
  11. 11 Kimi K2P6 5.00%
  12. 12 Grok 4.3 1.59%

Source: The scBench-Long (original study) paper

#EntryScore ±SettingDateSource
1 Claude Opus 4.8 25.4% 16.3 to 37.3 Claude Code May 2026 Official leaderboard
2 Gemini 3.5 Flash 22.2% 13.7 to 33.9 PI May 2026 Official leaderboard
3 Gemini 3.1 Pro 17.5% 10 to 28.6 PI Feb 2026 Official leaderboard
4 GPT-5.5 15.9% 8.9 to 26.8 OpenAI Codex Apr 2026 Official leaderboard
5 Claude Opus 4.7 11.1% 5.5 to 21.2 Claude Code Apr 2026 Official leaderboard
6 GLM-5P2 9.52% 4.4 to 19.3 PI Official leaderboard
7 Claude Opus 4.6 9.52% 4.4 to 19.3 PI Feb 2026 Official leaderboard
8 GPT-5.4 9.52% 4.4 to 19.3 PI Mar 2026 Official leaderboard
9 Claude Sonnet 4.6 7.94% 3.4 to 17.3 PI Feb 2026 Official leaderboard
10 Grok 4.20 6.35% 2.5 to 15.2 PI Feb 2026 Official leaderboard
11 Kimi K2P6 5.00% 1.7 to 13.7 PI Official leaderboard
12 Grok 4.3 1.59% 0.3 to 8.5 PI Apr 2026 Official leaderboard
#ModelScore ±SettingDateSource
1 Claude Opus 4.8 25.4% 16.3 to 37.3 Claude Code May 2026 Official leaderboard
2 Gemini 3.5 Flash 22.2% 13.7 to 33.9 PI May 2026 Official leaderboard
3 Gemini 3.1 Pro 17.5% 10 to 28.6 PI Feb 2026 Official leaderboard
4 GPT-5.5 15.9% 8.9 to 26.8 OpenAI Codex Apr 2026 Official leaderboard
5 Claude Opus 4.7 11.1% 5.5 to 21.2 Claude Code Apr 2026 Official leaderboard
6 Claude Opus 4.6 9.52% 4.4 to 19.3 PI Feb 2026 Official leaderboard
7 GPT-5.4 9.52% 4.4 to 19.3 PI Mar 2026 Official leaderboard
8 Claude Sonnet 4.6 7.94% 3.4 to 17.3 PI Feb 2026 Official leaderboard
9 Grok 4.20 6.35% 2.5 to 15.2 PI Feb 2026 Official leaderboard
10 Grok 4.3 1.59% 0.3 to 8.5 PI Apr 2026 Official leaderboard

Number one over time

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

Anthropic
Benchmark paper Not tested since Jun 2026

Source: The scBench-Long (original study) 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% 2026 Near the ceiling (85%) Claude Opus 4.8 Gemini 3.1 Pro Claude Opus 4.6 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Opus 4.8 Gemini 3.1 Pro

Source: Official scBench-Long (original study) 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

More in Biology and chemistry

All 137 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%
  • MedChemBench (OpenAI) In use Models work through 99 medicinal chemistry problems drawn from recent papers and patents. Each task mixes prose with structures, assay tables and plots, then asks for the program’s next decision. GPT-6 Astra 83.8%
  • 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%
  • 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%

Data audit

All 9 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access Public tasks, plus a private test set.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the benchmark’s paper (12 entries).
  • Pass Models scored 10 models we track, plus 2 other entries on its leaderboard.
  • Note Newest score The newest came 6 months ago, in May 2026, for Claude Opus 4.8. That’s expected for a test replaced by scBench-Long.
  • Note Frontier models None of the top 10 has a score, which is fine for a retired test. The strongest tested is GPT-5.5 (#21).
  • Pass Headroom The best is 25.4%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 15 rows for models we track and 2 other entries at the last read.
  • Pass Last read November 21st, 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