bix-35-q1 | 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

Calculate the evolutionary rate for the BUSCO gene 156083at2759 using PhyKIT's evoluionary_rate function. What is the gene's evolutionary rate in animals?

Accepted answer: 0.0471
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 3.93M total	3/3 replicate 1 replicate 2 replicate 3 1.46M total
Codex GPT-5.6 Sol	3/3 replicate 1 replicate 2 replicate 3 5.56M total	3/3 replicate 1 replicate 2 replicate 3 960.3K total
Codex GPT-5.6 Luna	3/3 replicate 1 replicate 2 replicate 3 6.79M total	3/3 replicate 1 replicate 2 replicate 3 3.16M total
DeepSeek V4 Pro via Codex	3/3 replicate 1 replicate 2 replicate 3 5.20M total	3/3 replicate 1 replicate 2 replicate 3 1.38M total
DeepSeek V4 Pro via Claude Codesuperseded	2/3 replicate 1 replicate 2 replicate 3 3.89M total	3/3 replicate 1 replicate 2 replicate 3 1.64M total

Execution condition

Galaxy-API code with skills

Codex GPT-5.5

3 primary replicates

Codex GPT-5.5

replicate 1

pass

Answer0.0471

Workflow

Phykit MetricsMAFFT

Total tokens: 1.73M
Output: 6.8K

Galaxy history Run trace

Codex GPT-5.5

replicate 2

pass

Answer0.0471

Workflow

Phykit MetricsMAFFTBUSCONewick analysis

Total tokens: 929.1K
Output: 8.3K

Galaxy history Run trace

Codex GPT-5.5

replicate 3

pass

Answer0.0471

Workflow

Phykit Tree BasedBUSCOMAFFTNewick analysis

Total tokens: 1.27M
Output: 5.3K

Galaxy history Run trace

Codex GPT-5.6 Sol

3 primary replicates

Codex GPT-5.6 Sol

replicate 1

pass

Answer0.0471

Workflow

Phykit Metrics

Total tokens: 1.49M
Output: 6.7K

Galaxy history Run trace

Codex GPT-5.6 Sol

replicate 2

pass

Answer0.0471

Workflow

Phykit Metrics

Total tokens: 1.03M
Output: 6.1K

Galaxy history Run trace

Codex GPT-5.6 Sol

replicate 3

pass

Answer0.0471

Workflow

Phykit Metrics

Total tokens: 3.04M
Output: 8.3K

Galaxy history Run trace

Codex GPT-5.6 Luna

3 primary replicates

Codex GPT-5.6 Luna

replicate 1

pass

Answer0.0471

Workflow

Phykit MetricsBUSCOMAFFT

Total tokens: 2.64M
Output: 15.3K

Galaxy history Run trace

Codex GPT-5.6 Luna

replicate 2

pass

Answer0.0471

Workflow

Phykit MetricsBUSCOMAFFT

Total tokens: 1.91M
Output: 12.1K

Galaxy history Run trace

Codex GPT-5.6 Luna

replicate 3

pass

Answer0.0471

Workflow

Phykit MetricsBUSCO

Total tokens: 2.25M
Output: 14.4K

Galaxy history Run trace

DeepSeek V4 Pro via Codex

3 primary replicates

DeepSeek V4 Pro via Codex

replicate 1

pass

Answer0.0471

Workflow

Phykit MetricsMAFFTIQ-TREE

Total tokens: 3.51M
Output: 16.5K

Galaxy history Run trace

DeepSeek V4 Pro via Codex

replicate 2

pass

Answer0.0471

Workflow

Phykit MetricsMAFFT

Total tokens: 1.11M
Output: 13.5K

Galaxy history Run trace

DeepSeek V4 Pro via Codex

replicate 3

pass

Answer0.0471

Workflow

Phykit MetricsBUSCOMAFFT

Total tokens: 592.2K
Output: 9.8K

Galaxy history Run trace

DeepSeek V4 Pro via Claude CodeSuperseded

3 primary replicates

DeepSeek V4 Pro via Claude Code

replicate 1

fail

Why it missed

Wrong scientific quantityThis run returned total tree length, 0.1884. PhyKIT evolutionary_rate normalizes tree length by the number of terminals, giving the requested 0.0471.

Answer0.1884

Workflow

Phykit MetricsBUSCOMAFFTNewick analysis

Total tokens: 1.03M
Output: 11.8K

Galaxy history Run trace

DeepSeek V4 Pro via Claude Code

replicate 2

pass

Answer0.0471

Workflow

Phykit MetricsUnzipMAFFTNewick analysis

Total tokens: 1.76M
Output: 18.3K

Galaxy history Run trace

DeepSeek V4 Pro via Claude Code

replicate 3

pass

Answer0.0471

Workflow

Phykit MetricsBUSCOMAFFTNewick analysis

Total tokens: 1.11M
Output: 9.0K

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

Answer0.0471

Workflow

Local analysisPhyKITMAFFTIQ-TREE

Total tokens: 627.5K
Output: 3.2K

Run trace

Codex GPT-5.5

replicate 2

pass

Answer0.0471

Workflow

Local analysisPhyKITMAFFTIQ-TREEBUSCO

Total tokens: 320.1K
Output: 2.8K

Run trace

Codex GPT-5.5

replicate 3

pass

Answer0.0471

Workflow

Local analysisBUSCOPhyKITMAFFTIQ-TREE

Total tokens: 508.0K
Output: 4.4K

Run trace

Codex GPT-5.6 Sol

3 primary replicates

Codex GPT-5.6 Sol

replicate 1

pass

Answer0.0471

Workflow

Local analysis

Total tokens: 299.0K
Output: 2.4K

Run trace

Codex GPT-5.6 Sol

replicate 2

pass

Answer0.0471

Workflow

Local analysis

Total tokens: 299.4K
Output: 2.4K

Run trace

Codex GPT-5.6 Sol

replicate 3

pass

Answer0.0471

Workflow

Local analysis

Total tokens: 362.0K
Output: 2.4K

Run trace

Codex GPT-5.6 Luna

3 primary replicates

Codex GPT-5.6 Luna

replicate 1

pass

Answer0.0471

Workflow

Local analysis

Total tokens: 1.26M
Output: 9.1K

Run trace

Codex GPT-5.6 Luna

replicate 2

pass

Answer0.0471

Workflow

Local analysis

Total tokens: 1.12M
Output: 7.2K

Run trace

Codex GPT-5.6 Luna

replicate 3

pass

Answer0.0471

Workflow

Local analysis

Total tokens: 772.9K
Output: 8.3K

Run trace

DeepSeek V4 Pro via Codex

3 primary replicates

DeepSeek V4 Pro via Codex

replicate 1

pass

Answer0.0471

Workflow

Local analysisBUSCOPhyKITMAFFT

Total tokens: 359.2K
Output: 4.8K

Run trace

DeepSeek V4 Pro via Codex

replicate 2

pass

Answer0.0471

Workflow

Local analysisPhyKITBUSCOMAFFTIQ-TREE

Total tokens: 570.6K
Output: 6.3K

Run trace

DeepSeek V4 Pro via Codex

replicate 3

pass

Answer0.0471

Workflow

Local analysisBUSCOPhyKITMAFFT

Total tokens: 445.3K
Output: 5.7K

Run trace

DeepSeek V4 Pro via Claude CodeSuperseded

3 primary replicates

DeepSeek V4 Pro via Claude Code

replicate 1

pass

Answer0.0471

Workflow

Local analysisPhyKITBUSCOMAFFT

Total tokens: 371.9K
Output: 3.1K

Run trace

DeepSeek V4 Pro via Claude Code

replicate 2

pass

Answer0.0471

Workflow

Local analysisPhyKITBUSCOMAFFT

Total tokens: 582.7K
Output: 4.2K

Run trace

DeepSeek V4 Pro via Claude Code

replicate 3

pass

Answer0.0471

Workflow

Local analysisBUSCOPhyKITMAFFT

Total tokens: 686.1K
Output: 4.6K

Run trace

bix-35-q1 · 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

Answer0.0471

Total tokens: 610.6K
Input: 606.5K
Cached input: 568.6K
Output: 4.1K

Run trace Final answer Galaxy history

Galaxy-API code · October 2026

replicate 2

Correct

Answer0.0471

Total tokens: 690.3K
Input: 684.9K
Cached input: 646.1K
Output: 5.5K

Run trace Final answer Galaxy history

Galaxy-API code · October 2026

replicate 3

Correct

Answer0.0471

Total tokens: 800.2K
Input: 795.2K
Cached input: 746.2K
Output: 5.1K

Run trace Final answer Galaxy history

Open-ended code averaged 0.32M tokens; previous website Galaxy averaged 1.85M; the new Galaxy runs averaged 0.70M, or 2.19× 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