DataLearner logo

Gemini 3.6 FlashvsGemini 3.0 Flash

Across 4 shared benchmarks, Gemini 3.6 Flash leads overall: Gemini 3.6 Flash wins 3, Gemini 3.0 Flash wins 1, with 0 ties and an average score difference of +6.83.

Google Deep Mind
Gemini 3.6 Flash

Google Deep Mind · 2026-07-21 · Multimodal model

Google Deep Mind
Gemini 3.0 Flash

Google Deep Mind · 2025-12-17 · Chat model

Gemini 3.6 Flash3 wins(75%)(25%)1 winGemini 3.0 Flash

Benchmark scores

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

Coding and Software Engineer

Even 2/2
BenchmarkGemini 3.6 FlashGemini 3.0 FlashDiff
SWE-Bench Pro - Public58.7012 / 55Thinking (With Tools)49.6043 / 55Thinking High (With Tools)+9.10
WeirdML v256.1033 / 52Thinking High (With Tools)61.6026 / 52Normal (With Tools)-5.50

AI Agent - Tool Usage

Gemini 3.6 Flash 1/1
BenchmarkGemini 3.6 FlashGemini 3.0 FlashDiff
Terminal-Bench 2.17817 / 37Thinking (With Tools)5834 / 37Thinking High (With Tools)+20

General Evaluation

Gemini 3.6 Flash 1/1
BenchmarkGemini 3.6 FlashGemini 3.0 FlashDiff
GPQA Diamond94.136 / 225Thinking High (No Tools)90.4036 / 225+3.73

Specs

FieldGemini 3.6 FlashGemini 3.0 Flash
PublisherGoogle Deep MindGoogle Deep Mind
Release date2026-07-212025-12-17
Model typeMultimodal modelChat model
ArchitectureDenseDense
ParametersNot availableNot available
Context length1M2000K
Max output64K64K

API pricing

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

ItemGemini 3.6 FlashGemini 3.0 Flash
Text input$1.5 / 1M tokens$0.5 / 1M tokens
Text output$7.5 / 1M tokens$3 / 1M tokens
Cache read$0.15 / 1M tokensNot public

Summary

  • Gemini 3.6 Flashleads in:AI Agent - Tool Usage (1/1), General Evaluation (1/1)
  • Tied in:Coding and Software Engineer

On average across the 4 shared benchmarks, Gemini 3.6 Flash scores 6.83 higher.

Largest single-benchmark gap: Terminal-Bench 2.1 — Gemini 3.6 Flash 78 vs Gemini 3.0 Flash 58 (+20).

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