Breaking

Google releases Gemini 3.7 Flash for coding and agents, three weeks after 3.6 Flash

September 1, 2026 at 17:04 EDT

Foundation Models
AI Agents
Software Dev & Coding

Google released Gemini 3.7 Flash on August 13, 2026, calling it the company's most intelligent workhorse model yet for coding and agents and shipping it just three weeks after Gemini 3.6 Flash. The generally available model is offered through December 31 at an introductory API price of $0.75 per million input tokens and $3.75 per million output tokens, which Google describes as half the original 3.6 Flash cost per million tokens.

August 13, 2026 · Google

Gemini 3.7 Flash, built for coding agents

Google shipped its most capable Flash-tier model three weeks after 3.6 Flash, with higher software-engineering scores and introductory API pricing of $0.75 per million input tokens through December 31, 2026.

Long-horizon coding

DeepSWE v1.1

3.7 Flash closes much of the gap with 3.6 Flash. GPT-5.6 Terra still leads.

48.6%

3.6 Flash

65.3%

3.7 Flash

69.6%

GPT-5.6 Terra

3 wks

After 3.6 Flash

$0.75

Per 1M input tokens

43.6%

FrontierCode 1.1 Main

Input context window

API input price

Introductory rate is half the 2027 list price

$0.75

Intro, through Dec 31

$1.50

From Jan 1, 2027

Output follows the same split: $3.75 now, $7.50 after the promo window.

Where it helps

Higher first-pass coding than 3.6 Flash, cheaper agent runs in early tests, and Spark tool use across Workspace apps.

Where it still lags

GPT-5.6 Terra still leads DeepSWE. A new Pro flagship has not shipped. Token prices double on January 1, 2027.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.