Who leads the BixBench benchmark?

GPT-4o currently leads BixBench with a score of 17 on Accuracy.

What is the state-of-the-art score on BixBench?

The state-of-the-art result on BixBench is 17 (Accuracy), achieved by GPT-4o.

How many models are tracked on BixBench?

Codesota tracks 2 models on BixBench.

Codesota · Benchmark · BixBenchHome/Leaderboards/Science, Medicine & Industry/Bioinformatics Agents/BixBench

Unknown

BixBench.

Name: BixBench Benchmark Results
Creator: Unknown
License: https://creativecommons.org/licenses/by/4.0/

50+ real-world biological data-analysis scenarios with ~300 open-answer questions designed to measure LLM agents on long, multi-step analytical trajectories.

Paper ↗Leaderboard ↓

§ 01 · Leaderboard

Results by metric.

Only 2 models on this benchmark

Help build the community leaderboard — submit your model results.

Found a wrong score or missing run?

Use row edits to send a sourced correction into moderation.

Add / edit result ↗Report issue ↗

Accuracy

Accuracy is the reported evaluation metric for BixBench. Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.

Higher is better

Trust tiers for Accuracyverifiedpapervendorcommunityunverified

Rank	Model	Trust	Score	Year	Links	Fix
01	GPT-4o Open-answer accuracy. From BixBench paper abstract: frontier models achieve ~17%. Per-model number not yet cross-checked against table in paper body.	paper	17	N/A	Paper ↗	Looks wrong?
02	Claude 3.5 Sonnet Open-answer accuracy. From BixBench paper abstract: frontier models achieve ~17%. Per-model number not yet cross-checked against table in paper body.	paper	17	N/A	Paper ↗	Looks wrong?

§ 04 · Submit a result

Add to the leaderboard.

← Back to Bioinformatics Agents