Updated

Luminous light ribbons sweeping diagonally through three small glass prisms on a dark background, abstract hero artwork for Gemini 3.7 Flash

Google

Gemini 3.7 Flash

Google's August 2026 Flash model for coding and agents, built to get code right on the first pass.

Chat with Gemini 3.7 Flash in Super AgentView pricing

What is Gemini 3.7 Flash?

Gemini 3.7 Flash is a Google model released on August 13, 2026, which Google called its most intelligent workhorse model yet for coding and agents.

Google said it writes more accurate, production-ready code on the first pass than Gemini 3.6 Flash, builds more complete layouts and apps from a screenshot or design system, and adds agentic video understanding. Super Agent runs it on Google Cloud's Vertex AI, where Google plans to retire it on January 28, 2027 and recommends Gemini 3.8 Flash as the upgrade.

Provider
Google
Context window
1M tokens
Input
Text, Image, Video, Audio, PDF
Reasoning
Low, medium, high
Max output
65,536 tokens
Knowledge cutoff
March 2026
Released
Aug 13, 2026

How Gemini 3.7 Flash performs

Published by Google

Results Google reported for Gemini 3.7 Flash. Each lab runs its own setup, so compare these with other models only inside the same table.

Benchmark results Google published for Gemini 3.7 Flash
BenchmarkScoreSetting
FrontierCode 1.1 Main43.6%
DeepSWE v1.165.3%
Terminal-Bench 2.185.8%
AutomationBench30.4%private set
GDP.pdf34%
CharXiv Reasoning84.5%no tools
LVBench85.4%
GDM-MRCR v2, 8 needles97%128K, average
OSWorld 2.047.9%
HLE-Verified53.6%

Source: deepmind.google

What you can do with Gemini 3.7 Flash

  • Higher first-pass code accuracy than Gemini 3.6 Flash, according to Google

  • Builds layouts and apps that follow a screenshot or design system

  • Agentic video understanding, which Google also added to 3.6 Flash and 3.5 Flash-Lite

  • Reads text, images, video, audio and PDFs

  • 1M-token context window

How to prompt Gemini 3.7 Flash

  • Keep it on low effort for speed and raise it to high for multi-step plans.
  • Attach the screenshot or design you want a layout to match.
  • Ask for numbered steps when you want it to reason through a problem.

Where Gemini 3.7 Flash fits best

  • Fast agent runs

  • Front-end code from a screenshot or mockup

  • Questions about a video clip

  • Knowledge-heavy questions in finance, law and biosciences

How Gemini 3.7 Flash compares

Gemini 3.7 Flash compared with other chat models on Flashloop
ModelLMArena textLMArena codingContext windowReasoning
Gemini 3.7 Flash1490#161520#361M tokensLow, medium, high
Claude Opus 51491#131528#231M tokensOff to max
Kimi K31487#181539#111M tokensOff, low, medium, high, max
Gemini 3.1 Pro Preview1487#191522#321M tokensLow, medium, high
GPT 5.6 Sol1485#211534#171.05M tokensOff to max

Good to know before you start

  • Google plans to retire it on Vertex AI on January 28, 2027
  • Three reasoning levels; no off or max setting
  • Reads video and audio but writes only text; Flashloop's generation models create the media

Start chatting with Gemini 3.7 Flash

Chat with Gemini 3.7 Flash in Super Agent

  1. Select Gemini 3.7 Flash

    Open Super Agent on Flashloop and choose Gemini 3.7 Flash in the model picker.

  2. Enter your prompt

    Ask for what you need: an answer, a plan, or the images and videos you want the agent to make.

  3. Send and keep going

    Send the message. Gemini 3.7 Flash answers in Super Agent and runs any generations you asked for.

Questions about Gemini 3.7 Flash

Which model answers when I pick Gemini 3.7 Flash?

Gemini 3.7 Flash itself. Super Agent calls it through Google Cloud's Vertex AI. On October 8, 2026 Google started sending requests for this name to Gemini 3.8 Flash on the Gemini API, but Vertex AI still serves Gemini 3.7 Flash and lists its retirement for January 28, 2027.

What was Gemini 3.7 Flash designed for?

Google built it as a workhorse for coding and agents, with gains in first-pass code accuracy, layouts that follow a design, knowledge-heavy fields and video understanding.

What inputs does Gemini 3.7 Flash accept?

Text, images, video, audio and PDFs, with a 1M-token context window.

Which reasoning levels does Gemini 3.7 Flash support in Super Agent?

Low, medium and high. Super Agent starts it at low.

More chat models on Flashloop

Browse all models