Model result, rank 15 of 27

DeepSeek V4 Pro

DeepSeek direct API · maximum-reasoning Hard IQ. 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 #11 of 27

80.06

SWE MVPlane rank #11 of 27

79.68

Hard Intelligencelane rank #16 of 27

78.38

DeepSeek V4 Pro
Cohort median
Cohort best

Overall score79.37

Rank #15

Full / Agentic80.06

Full rank #11

SWE MVP79.68

SWE rank #11

Hard Intelligence78.38

Hard rank #16

Measured cost$0.335

100.0% reliability

Overall

All-around publication view

Score79.37

Formulamean(Full, SWE, Hard Intelligence)

BasisDeepSeek direct API · maximum-reasoning Hard IQ

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

Lane 01

Full / Agentic benchmark

Final80.06

Capability93.50

Agentic86.69

Pass rate88.6%

Prompts43

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

Lane 02

Software engineering MVP

SWE score79.68

Focused final64.26

Capability75.75

Daily driver60.42

Prompts24

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

Lane 03

Hard Intelligence diagnostic

Hard score78.38

Active inquiry81.42

Online adaptation80.00

Self-repair82.08

Authority integrity70.00

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

Telemetry

Runtime economics

Total cost$0.335

Cost / scored item$0.0045

Seconds / timed item23.25s

Runtime coverage100.0%

Recorded tokens / item9.6k

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 78.38 and contributes to the overall score alongside Full/Agentic and SWE. Hard Intelligence is published alongside Full/Agentic and SWE for the current ranking.