DeepSeek V4 Flash 0731
DeepSeek · 1 configuration
Intelligence34.3
Input / 1M$0.44
Output / 1M$1.32
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 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows....
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
34.3 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
90.8 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
38.6 %SciCodeCode generation for scientific research problems.
50.3 %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.
40.4 %Specifications from the catalog
Inputtext
Outputtext
Tool callingYes
Structured outputYes
Speed218.0 tok/s
First token1.04 s
Maximum output943.7K
Catalog listing date2026-07-31