BenchLeader

Gemini 3.6 Flash vs Gemini 3.7 Flash

Verdict
  • Gemini 3.7 Flash (medium) leads on quality: 64.5 vs 60.8.
  • Gemini 3.6 Flash is stronger in maths.
  • Gemini 3.7 Flash (medium) is stronger in coding, composite, knowledge, long context, multimodal, reasoning.
  • They cost about the same ($1.50 per 1M blended).
  • Gemini 3.7 Flash (medium) streams 1.4× faster (272 vs 194 tokens per second).
MetricGemini 3.6 FlashGemini 3.7 Flash (medium)
BenchLeader Index60.864.5
Coding score49.868.5
Composite score72.479.1
Knowledge score74.775.4
Long context score66.668.1
Maths score54.8
Multimodal score67.869.4
Reasoning score63.5
Blended price $/M$1.50$1.50
Output speed194 tok/s272 tok/s
Time to first answer16.8 s4.6 s
Context window1.0M1M
SciCode57.9%
FrontierCode34.4%
ProofBench36.0%
Epoch Capabilities Index154.3
AA Intelligence Index34.339.6
AA-LCR80.0%83.0%
MMMU-Pro83.2%84.7%
AA-Omniscience22.123.7
GPQA Diamond (AA)92.8%92.1%
Humanity's Last Exam (AA)40.8%39.0%
SciCode (AA)53.4%59.8%
AIME 202696.7%
HMMT February 202689.4%
MathArena Apex26.0%
ARC-AGI-191.2%
ARC-AGI-263.8%

Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.

Gemini 3.6 Flash vs Gemini 3.7 Flash: questions

Is Gemini 3.6 Flash better than Gemini 3.7 Flash?
Gemini 3.7 Flash (medium) leads on quality: 64.5 vs 60.8. The BenchLeader Index combines every independent quality benchmark; Gemini 3.7 Flash (medium) is ahead overall as of 2026-09-10, but check the category scores for your use.
Is Gemini 3.6 Flash better than Gemini 3.7 Flash for coding?
Gemini 3.7 Flash scores higher in coding (69 vs 50 on the category index, where 50 is average).
Which is cheaper, Gemini 3.6 Flash or Gemini 3.7 Flash?
Gemini 3.6 Flash is cheaper: $1.50 against $1.50 per million tokens, blended at three input tokens per output token.
Which is faster, Gemini 3.6 Flash or Gemini 3.7 Flash?
Gemini 3.7 Flash streams faster: 272 against 194 output tokens per second.
Which has the larger context window?
Gemini 3.6 Flash accepts more context: 1.0M against 1M tokens.