DeepSeek V4 Pro
DeepSeek · 3 thinking configurations
Intelligence20.8
Input / 1M$0.44
Output / 1M$0.87
Context1M
Tariff: DeepSeek · Source: artificial-analysis · USD per million tokens
Thinking configuration
What does more thinking buy you?
Compare this model’s measured thinking configurations.
Scroll the chart ↔ · tap a point for details
Efficient frontier: best measured score at each cost1 measured configurations · 2 without both values
Selected: thinking none · Click a point to switch configuration
Open MoE flagship with million-token context for coding and long agent runs
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
20.8 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
71.7 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
8.2 %τ²-BenchTool use and interaction in assistance scenarios. Does not replace an evaluation of your integration.
91.2 %IFBenchVerifiable instruction following. A proxy for precision rather than a benchmark of creative style.
45.8 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
30.9 %Specifications from the catalog
Inputtext
Outputtext
Tool callingYes
Structured outputYes
Speed78.5 tok/s
First token1.67 s
Maximum output16.4K
Catalog listing date2026-04-24