Model result, rank 14 of 27

Claude Sonnet 5

OpenRouter · extra-high reasoning · Full + SWE + Hard measured. Public result card with the model’s overall score, lane measurements, runtime and cost telemetry, and the ranking formula.

Compare all models Download public JSON

Figure 1 Lane profile against the cohort Each bar is one measured lane on the shared 0 to 100 scale. The thin mark is the cohort median; the bright mark is the best score any ranked model reached on that lane.

Full / Agenticlane rank #18 of 27

76.72

SWE MVPlane rank #12 of 27

78.08

Hard Intelligencelane rank #12 of 27

86.53

Claude Sonnet 5
Cohort median
Cohort best

Overall score80.44

Rank #14

Full / Agentic76.72

Full rank #18

SWE MVP78.08

SWE rank #12

Hard Intelligence86.53

Hard rank #12

Measured cost$2.812

100.0% reliability

Overall

All-around publication view

Score80.44

Formulamean(Full, SWE, Hard Intelligence)

BasisOpenRouter · extra-high reasoning · Full + SWE + Hard measured

The overall score averages the measured major lanes while keeping each source measurement visible.

Lane 01

Full / Agentic benchmark

Final76.72

Capability88.48

Agentic79.27

Pass rate88.6%

Prompts43

This lane captures instruction following, structured behavior, tool discipline, and general agentic reliability.

Lane 02

Software engineering MVP

SWE score78.08

Focused final72.02

Capability75.05

Daily driver71.94

Prompts24

This lane is closer to implementation usefulness: source handling, architecture cleanliness, and deliverable quality.

Lane 03

Hard Intelligence diagnostic

Hard score86.53

Active inquiry98.50

Online adaptation99.06

Self-repair86.04

Authority integrity62.50

Hard Intelligence measures active inquiry, online adaptation, evidence-driven self-repair, and authority/salience integrity.

Telemetry

Runtime economics

Total cost$2.812

Cost / scored item$0.034

Seconds / timed item12.51s

Runtime coverage100.0%

Recorded tokens / item13.7k

Token coverage100.0%

Cost, time, and token basis are normalized telemetry. They explain tradeoffs; they do not overwrite the capability score yet.

Interpretation

Why the result lands here.

The result is best read as a balanced benchmark entry: one overall score plus the lane measurements that produced it. Hard Intelligence score is 86.53 and contributes to the overall score alongside Full/Agentic and SWE. Measured with OpenRouter extra-high reasoning across Full/Agentic, SWE MVP, and Hard Intelligence. Hard Intelligence is a public diagnostic, not a hidden official score. Full long-context probe rescored under suite v4 (rubric: correct answer zeroed by the phrase list).