Google Cuts Gemini 3.7 Flash Pricing in Half While Doubling Coding Benchmark Scores

News Summary

Google released Gemini 3.7 Flash on August 13, 2026, positioning it as the company's "most intelligent workhorse model yet for coding and agents." The launch comes just three weeks after Gemini 3.6 Flash, marking one of the fastest iteration cycles Google has shipped for a low-cost model tier, and it arrives ahead of the anticipated Gemini 3.5 Pro release.

What's New in Gemini 3.7 Flash

Gemini 3.7 Flash targets three core workflows: software engineering, knowledge work, and web development. Google says the model delivers substantial gains over its predecessor across these areas, with particular emphasis on autonomous coding tasks and multi-step agent workflows that require tool use and sustained reasoning.

The model is built to handle longer, more complex agentic sessions — chaining together code edits, file operations, and external tool calls with less human intervention than previous Flash-tier models required.

Benchmark Performance

Google published a range of benchmark results comparing Gemini 3.7 Flash to Gemini 3.6 Flash, at 10:00 AM Eastern Time on the day of the announcement:

• FrontierCode 1.1 Main: 43.6%, up from 34.4%
• DeepSWE v1.1: 65.3%, up from 49.0%
• GDP.pdf (complex document processing): 34.0%, up from 22.0%
• AutomationBench (real-world business workflow completion): 30.4%, up from 17.0%
• WebDev Arena Elo rating: 1588, up from 1538

These gains span coding accuracy, document understanding, and the model's ability to complete practical, multi-step business tasks — three areas Google has prioritized as enterprise adoption of AI agents accelerates.

Pricing and Availability

Gemini 3.7 Flash launches with an introductory price of $0.75 per million input tokens and $3.75 per million output tokens, roughly half the initial launch price of Gemini 3.6 Flash. Google says this promotional pricing holds through the end of 2026, after which rates rise to $1.50 per million input tokens and $7.50 per million output tokens starting January 1, 2027, Pacific Time.

The model is available immediately through the Gemini API, Google AI Studio, Antigravity, Google Cloud Console, Vertex AI, and Gemini Enterprise, giving developers and enterprise customers multiple entry points depending on their existing Google Cloud workflows.

Safety Safeguards

Alongside the performance and pricing updates, Google confirmed that Gemini 3.7 Flash ships with updated safeguards addressing misuse risks in Chemical, Biological, Radiological, and Nuclear (CBRN) domains as well as cyber offense scenarios. The company said these safeguards were evaluated and strengthened as part of the standard pre-release testing process applied to its Gemini model family.

Why It Matters

The rapid three-week turnaround between Gemini 3.6 Flash and Gemini 3.7 Flash signals an accelerating release cadence for Google's lower-cost model tier, which competes directly on price and speed for high-volume coding and agent workloads. By cutting introductory pricing in half while posting double-digit benchmark improvements, Google is aiming to make Gemini 3.7 Flash an attractive default for developers building autonomous coding assistants and business-process agents at scale, ahead of the heavier Gemini 3.5 Pro model expected later this year.