DeepSeek V4 Pro 0813
DeepSeek · 1 configuration
Intelligence36.0
Input / 1M$1.32
Output / 1M$3.96
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 · 0 without both values
Selected: thinking max · Click a point to switch configuration
DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro.
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
36.0 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
92.8 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
41.0 %SciCodeCode generation for scientific research problems.
51.0 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
78.7 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
49.1 %Specifications from the catalog
Inputtext
Outputtext
Tool callingYes
Structured outputYes
Speed81.0 tok/s
First token1.67 s
Maximum output393.2K
Catalog listing date2026-08-12