AI Hub
← All models

DeepSeek-V4-Flash vs Olmo 3 32B Think

DeepSeek vs Allen Institute for AI — benchmarks, pricing, and capabilities side by side.

  • DeepSeek-V4-Flash has the higher intelligence index (89.4 vs 69.5)
  • DeepSeek-V4-Flash is cheaper ($0.10 vs $0.15 per 1M input)
  • DeepSeek-V4-Flash has a larger context window (1M)
DeepSeek-V4-FlashOlmo 3 32B Think
Intelligence index89.469.5
DeveloperDeepSeekAllen Institute for AI
TypeLLMLLM
AccessOpen weightsOpen weights
Context window1,048,576 tokens65,536 tokens
Input price$0.10 / 1M$0.15 / 1M
Output price$0.20 / 1M$0.50 / 1M
Speed109 tok/s
ReleasedApril 24, 2026November 21, 2025
Parameters284B (13B active)
Input modalitiesTextText
Output modalitiesTextText

Shared benchmarks

DeepSeek-V4-Flash
Olmo 3 32B Think
GPQA Diamond
89.4
61
Humanity’s Last Exam
32.1
5.9