DataLearner logo

Gemini 3.7 FlashvsDeepSeek-V4-Flash

Across 4 shared benchmarks, Gemini 3.7 Flash leads overall: Gemini 3.7 Flash wins 4, DeepSeek-V4-Flash wins 0, with 0 ties and an average score difference of +5.10.

Google Deep Mind
Gemini 3.7 Flash

Google Deep Mind · 2026-08-13 · Multimodal model

DeepSeek-AI
DeepSeek-V4-Flash

DeepSeek-AI · 2026-04-24 · Reasoning model

Gemini 3.7 Flash4 wins(100%)(0%)0 winsDeepSeek-V4-Flash

Benchmark scores

Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.

AI Agent - Tool Usage

Gemini 3.7 Flash 2/2
BenchmarkGemini 3.7 FlashDeepSeek-V4-FlashDiff
AutomationBench30.404 / 8Thinking (With Tools)25.108 / 8Max (With Tools)+5.30
Terminal-Bench 2.185.809 / 44Thinking (With Tools)82.7016 / 44Max (With Tools)+3.10

Agent Level Benchmark

Gemini 3.7 Flash 1/1
BenchmarkGemini 3.7 FlashDeepSeek-V4-FlashDiff
Agents' Last Exam26.308 / 11Thinking Medium (With Tools)25.2010 / 11Max (With Tools)+1.10

Coding and Software Engineer

Gemini 3.7 Flash 1/1
BenchmarkGemini 3.7 FlashDeepSeek-V4-FlashDiff
DeepSWE65.3010 / 27Thinking High (With Tools)54.4015 / 27Max (With Tools)+10.90

Specs

FieldGemini 3.7 FlashDeepSeek-V4-Flash
PublisherGoogle Deep MindDeepSeek-AI
Release date2026-08-132026-04-24
Model typeMultimodal modelReasoning model
ArchitectureDenseMoE
ParametersNot available284B
Context length1M1M
Max output64K384K

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemGemini 3.7 FlashDeepSeek-V4-Flash
Text input$0.75 / 1M tokens$0.14 / 1M tokens
Text output$3.75 / 1M tokens$0.28 / 1M tokens
Cache read$0.075 / 1M tokens$0.0028 / 1M tokens

Summary

  • Gemini 3.7 Flashleads in:AI Agent - Tool Usage (2/2), Agent Level Benchmark (1/1), Coding and Software Engineer (1/1)

On average across the 4 shared benchmarks, Gemini 3.7 Flash scores 5.10 higher.

Largest single-benchmark gap: DeepSWE — Gemini 3.7 Flash 65.30 vs DeepSeek-V4-Flash 54.40 (+10.90).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.