bix-34-q2 | Current BixBench results

New token-optimization experiment

Three additional GPT-5.6 Sol Galaxy replicates are available separately. Previous results remain unchanged.

View the October 2026 batch and its run records

Question

What is the median patristic distance for the fungal gene 981902at2759?

Accepted answer: 2.63
Domain: phylogenetics

Comparison

Pass counts and token totals across the same three replicates.

Agent	Galaxy-API code	Open-ended code
Codex GPT-5.5	3/3 replicate 1 replicate 2 replicate 3 5.56M total	3/3 replicate 1 replicate 2 replicate 3 745.1K total
Codex GPT-5.6 Sol	3/3 replicate 1 replicate 2 replicate 3 2.78M total	3/3 replicate 1 replicate 2 replicate 3 636.5K total
Codex GPT-5.6 Luna	3/3 replicate 1 replicate 2 replicate 3 11.99M total	3/3 replicate 1 replicate 2 replicate 3 860.2K total
DeepSeek V4 Pro via Codex	3/3 replicate 1 replicate 2 replicate 3 4.79M total	3/3 replicate 1 replicate 2 replicate 3 4.52M total
DeepSeek V4 Pro via Claude Codesuperseded	3/3 replicate 1 replicate 2 replicate 3 6.32M total	3/3 replicate 1 replicate 2 replicate 3 2.18M total

Execution condition

Galaxy-API code with skills

Codex GPT-5.5

3 primary replicates

Codex GPT-5.5

replicate 1

passUDT-only

Answer2.6322860447

Workflow

Galaxy UDTNewick analysisPhyKITMAFFTIQ-TREEBiopython

Total tokens: 1.36M
Output: 7.9K

Galaxy history Run trace

Codex GPT-5.5

replicate 2

pass

Answer2.6323

Workflow

UnzipPhykit Tree BasedBUSCONewick analysisMAFFT

Total tokens: 2.32M
Output: 7.4K

Galaxy history Run trace

Codex GPT-5.5

replicate 3

passUDT + standard tools

Answer2.6322860447

Workflow

Phykit Tree BasedGalaxy UDTBUSCOMAFFTNewick analysisBiopython

Total tokens: 1.88M
Output: 12.3K

Galaxy history Run trace

Codex GPT-5.6 Sol

3 primary replicates

Codex GPT-5.6 Sol

replicate 1

pass

Answer2.6323

Workflow

Phykit Tree BasedMAFFTNewick analysis

Total tokens: 862.2K
Output: 5.0K

Galaxy history Run trace

Codex GPT-5.6 Sol

replicate 2

passUDT-only

Answer2.6322860447

Workflow

Galaxy UDTBiopythonMAFFTmedian summary

Total tokens: 968.6K
Output: 7.4K

Galaxy history Run trace

Codex GPT-5.6 Sol

replicate 3

passUDT-only

Answer2.6322860447

Workflow

Galaxy UDTBUSCOBiopythonMAFFTNewick analysismedian summary

Total tokens: 948.1K
Output: 6.9K

Galaxy history Run trace

Codex GPT-5.6 Luna

3 primary replicates

Codex GPT-5.6 Luna

replicate 1

passUDT-only

Answer2.6322860447

Workflow

Galaxy UDTMAFFTIQ-TREENewick analysismedian summary

Total tokens: 1.67M
Output: 18.4K

Galaxy history Run trace

Codex GPT-5.6 Luna

replicate 2

passUDT-only

Answer2.6322860446999998

Workflow

Galaxy UDTBiopythonMAFFTNewick analysismedian summary

Total tokens: 5.30M
Output: 14.8K

Galaxy history Run trace

Codex GPT-5.6 Luna

replicate 3

pass

Answer2.6323

Workflow

UnzipPhykit Tree BasedMAFFT

Total tokens: 5.02M
Output: 15.3K

Galaxy history Run trace

DeepSeek V4 Pro via Codex

3 primary replicates

DeepSeek V4 Pro via Codex

replicate 1

pass

Answer2.6323

Workflow

Phykit Tree BasedMAFFTNewick analysis

Total tokens: 1.23M
Output: 10.4K

Galaxy history Run trace

DeepSeek V4 Pro via Codex

replicate 2

pass

Answer2.6323

Workflow

Phykit Tree BasedBUSCOMAFFTNewick analysis

Total tokens: 1.37M
Output: 9.2K

Galaxy history Run trace

DeepSeek V4 Pro via Codex

replicate 3

pass

Answer2.6323

Workflow

Phykit Tree BasedMAFFTNewick analysis

Total tokens: 2.19M
Output: 13.8K

Galaxy history Run trace

DeepSeek V4 Pro via Claude CodeSuperseded

3 primary replicates

DeepSeek V4 Pro via Claude Code

replicate 1

passUDT-only

Answer2.6322860447

Workflow

Galaxy UDTBiopythonMAFFTNewick analysismedian summarymean summary

Total tokens: 1.45M
Output: 17.9K

Galaxy history Run trace

DeepSeek V4 Pro via Claude Code

replicate 2

pass

Answer2.63225

Workflow

UnzipPhykit Tree Based

Total tokens: 1.59M
Output: 11.9K

Galaxy history Run trace

DeepSeek V4 Pro via Claude Code

replicate 3

passUDT + standard tools

Answer2.6322860447

Workflow

UnzipGalaxy UDTNewick analysis

Total tokens: 3.28M
Output: 32.3K

Galaxy history Run trace

Execution condition

Open-ended code with skills

Codex GPT-5.5

3 primary replicates

Codex GPT-5.5

replicate 1

pass

Answer2.6322860447

Workflow

Local analysisBUSCOBiopythonMAFFTIQ-TREENewick analysis

Total tokens: 209.1K
Output: 2.7K

Run trace

Codex GPT-5.5

replicate 2

pass

Answer2.6322860447

Workflow

Local analysisBUSCOMAFFTIQ-TREENewick analysismedian summary

Total tokens: 245.5K
Output: 2.7K

Run trace

Codex GPT-5.5

replicate 3

pass

Answer2.6322860447

Workflow

Local analysisBUSCOMAFFTIQ-TREENewick analysismedian summary

Total tokens: 290.5K
Output: 3.2K

Run trace

Codex GPT-5.6 Sol

3 primary replicates

Codex GPT-5.6 Sol

replicate 1

pass

Answer2.6322860447

Workflow

Local analysis

Total tokens: 179.4K
Output: 2.4K

Run trace

Codex GPT-5.6 Sol

replicate 2

pass

Answer2.6322860447

Workflow

Local analysis

Total tokens: 243.5K
Output: 2.6K

Run trace

Codex GPT-5.6 Sol

replicate 3

pass

Answer2.6322860447

Workflow

Local analysis

Total tokens: 213.5K
Output: 2.3K

Run trace

Codex GPT-5.6 Luna

3 primary replicates

Codex GPT-5.6 Luna

replicate 1

pass

Answer2.6322860447

Workflow

Local analysis

Total tokens: 292.3K
Output: 4.8K

Run trace

Codex GPT-5.6 Luna

replicate 2

pass

Answer2.6322860447

Workflow

Local analysis

Total tokens: 284.6K
Output: 4.8K

Run trace

Codex GPT-5.6 Luna

replicate 3

pass

Answer2.6322860447

Workflow

Local analysis

Total tokens: 283.3K
Output: 4.6K

Run trace

DeepSeek V4 Pro via Codex

3 primary replicates

DeepSeek V4 Pro via Codex

replicate 1

pass

Answer2.6322860447

Workflow

Local analysisMAFFTBUSCOIQ-TREENewick analysismedian summary

Total tokens: 1.10M
Output: 16.5K

Run trace

DeepSeek V4 Pro via Codex

replicate 2

pass

Answer2.6322860447

Workflow

Local analysisBUSCOMAFFTIQ-TREEPhyKIT

Total tokens: 825.8K
Output: 9.0K

Run trace

DeepSeek V4 Pro via Codex

replicate 3

pass

Answer2.6323

Workflow

Local analysisBUSCOMAFFTIQ-TREEmedian summaryNewick analysis

Total tokens: 2.60M
Output: 22.0K

Run trace

DeepSeek V4 Pro via Claude CodeSuperseded

3 primary replicates

DeepSeek V4 Pro via Claude Code

replicate 1

pass

Answer2.6322860447

Workflow

Local analysisBUSCOMAFFTNewick analysismedian summaryIQ-TREE

Total tokens: 1.09M
Output: 19.5K

Run trace

DeepSeek V4 Pro via Claude Code

replicate 2

pass

Answer2.6322860447

Workflow

Local analysisBUSCOMAFFTIQ-TREENewick analysis

Total tokens: 488.5K
Output: 6.4K

Run trace

DeepSeek V4 Pro via Claude Code

replicate 3

pass

Answer2.6322860447

Workflow

Local analysisBUSCOMAFFTNewick analysisIQ-TREE

Total tokens: 602.5K
Output: 16.6K

Run trace

bix-34-q2 · October 2026 · three additional replicates

GPT-5.6 Sol · token optimization

Galaxy-API code, high reasoning, Fast service. Previous runs are retained above.

Galaxy-API code · October 2026

replicate 1

Correct

Answer2.6322860447

Total tokens: 925.9K
Input: 914.7K
Cached input: 871.7K
Output: 11.2K

Run trace Final answer Galaxy history

Galaxy-API code · October 2026

replicate 2

Correct

Answer2.6322860447

Total tokens: 1.85M
Input: 1.83M
Cached input: 1.77M
Output: 14.5K

Run trace Final answer Galaxy history

Galaxy-API code · October 2026

replicate 3

Correct

Answer2.6322860447

Total tokens: 819.0K
Input: 810.8K
Cached input: 766.7K
Output: 8.2K

Run trace Final answer Galaxy history

Open-ended code averaged 0.21M tokens; previous website Galaxy averaged 0.93M; the new Galaxy runs averaged 1.20M, or 5.64× Open-ended code. Each mean covers three replicates of this same task with GPT-5.6 Sol.

Cached input is already included in input tokens. Full nested tool/job evidence is in each HF run's galaxy_execution_records.tar.gz, with an accompanying member index.

Optimization overview and all 150 runs