AI Hub
← All models

DeepSeek-V2.5 vs DeepSeek-V4-Flash

DeepSeek vs DeepSeek — benchmarks, pricing, and capabilities side by side.

  • DeepSeek-V4-Flash has the higher intelligence index (89.4 vs 63.4)
  • DeepSeek-V4-Flash is cheaper ($0.10 vs $0.14 per 1M input)
  • DeepSeek-V4-Flash is faster
  • DeepSeek-V4-Flash has a larger context window (1M)
DeepSeek-V2.5DeepSeek-V4-Flash
Intelligence index63.489.4
DeveloperDeepSeekDeepSeek
TypeLLMLLM
AccessOpen weightsOpen weights
Context window8,192 tokens1,048,576 tokens
Input price$0.14 / 1M$0.10 / 1M
Output price$0.28 / 1M$0.20 / 1M
Speed100 tok/s109 tok/s
ReleasedMay 8, 2024April 24, 2026
Parameters236000000000284B (13B active)
Input modalitiesText
Output modalitiesText