llm-benchmarks
githubdrose.io

google Provider Benchmarks


Measured over the last 30 days

Models Tracked
21
Avg Tokens / Second
35.34
Avg Time to First Token (ms)
3914.76
Last Updated
Oct 14, 2026

All google models

21 live · one row per model · fastest first · 3 retired lanes not listed · 4 folded behind their current lane
google site
ModelAvg Toks/SecMinMaxAvg TTF (ms)
Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)97.4075.40119.00710.00
Gemini 3.1 Flash Lite74.8062.8086.80680.00
Nano Banana 2 (Gemini 3.1 Flash Image Preview)72.0066.3077.60940.00
Nano Banana 2 (Gemini 3.1 Flash Image)69.6068.9070.201020.00
Gemini 3.5 Flash Lite67.6064.9070.30670.00
Gemini 2.5 Flash Lite66.5059.3073.70820.00
Gemini 2.5 Flash62.5060.5064.50650.00
Gemini 3.1 Flash Lite Preview61.3059.4063.30810.00
Gemini 3 Flash Preview46.0045.1047.00980.00
Lyria 3 Clip Preview34.1034.1034.103720.00
Nano Banana (Gemini 2.5 Flash Image)24.809.1740.401040.00
Gemini 4 Argon16.3016.3016.303400.00
Gemini 3.7 Flash12.1012.0012.206080.00
Gemini 3.8 Flash9.053.3116.208450.00
Gemini 3.6 Flash8.047.798.299140.00
Gemini Flash Latest5.614.427.5711300.00
Gemini 3.5 Flash5.415.255.583600.00
Gemini 2.5 Pro Preview 06-053.202.913.495950.00
Gemini 3.1 Pro Preview Custom Tools2.101.632.567320.00
Gemini 3.1 Pro Preview2.011.122.908430.00
Gemini 2.5 Pro1.711.711.716500.00

Frequently Asked Questions

Which google model is fastest?
Based on recent tests, Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) shows the highest average throughput among tracked google models.
How many recent measurements feed this dashboard?
This provider summary aggregates 219 individual prompts measured across 212 monitoring runs over the past month.