Community

Platform

Docs Roadmap Tools Migrate Pricing Model Pricing Calculator Works With

Support

Contribute Database monitor Status Contact

Company

About Announcements Changelog Terms Privacy

Community

Platform

Docs Roadmap Tools Migrate Pricing Model Pricing Calculator Works With

Support

Contribute Database monitor Status Contact

Company

About Announcements Changelog Terms Privacy

© 2025 AI StatsIf you run into any issues or notice any data errors, please visit our GitHub and report an issue.

Quickstart

Start calling this model with endpoint-specific examples.

Benchmarks

Headline benchmark standings and comparison context.

About

Key dates, capabilities, and model metadata.

Models Providers Apps Rankings Playground

Models Providers Apps Rankings Playground

DeepSeek V4 Flash Benchmarks - Performance Highlights | AI Stats

DeepSeek V4 Flash

DeepSeek V4 Flash

DeepSeek

Overview Playground Providers Pricing Performance Apps Activity Quickstart Benchmarks Family Timeline

Benchmarks

Headline benchmark standings and comparison context.

Highlights

Top benchmark results for deepseek/deepseek-v4-flash.

82.60-

33%-

85.70%-

86.90%-

56.80%-

73.20%-

92.10%-

Chinese SimpleQA

78.90%-

82.20%-

93.60%-

90.40-

3052-

60.50%-

88.60%-

FACTS Parametric

33.90%-

1395-

88.10%-

90.80%-

85.70%-

94.80%-

69.50%-

Humanity's Last Exam

45.10%-

IMO Answer Bench

88.40%-

91.60%-

44.70%-

57.40%-

85.70%-

88.70%-

89.40%-

86.40%-

88.80%-

78.70%-

42.20%-

Scale MCP Atlas

69%-

SimpleQA Verified

34.10-

46.50%-

SWE Bench Multilingual

73.30%-

52.60%-

79%-

Terminal Bench 2.0

56.90%-

47.80%-

82.80%-

79.50%-

Detailed benchmark comparisons now live in the Compare tool.

Open Compare Tool Browse Comparisons