Google releases Gemini 3.7 Flash at half the price of predecessor

👁 119 reads

Google has launched Gemini 3.7 Flash today, a replacement for 3.6 Flash just three weeks after that model shipped, pricing it at $0.75 per million input tokens and $3.75 per million output tokens - exactly half what 3.6 Flash costs - as an introductory rate available through the end of the year.

According to Senior Director Tulsee Doshi, the new model delivers meaningful coding gains. The FrontierCode 1.1 Main benchmark rose from 34.4% to 43.6%, and DeepSWE v1.1 climbed from 49% to 65.3%. The WebDev Arena score moved to 1,588 from 1,538. Document and workflow performance also improved: the GDP.pdf benchmark, which measures complex document processing, jumped from 22% to 34%, while AutomationBench, which tests common business workflow execution, rose from 17% to 30.4%.

Google describes 3.7 Flash as the product of core optimizations and developer feedback. The price cut appears aimed directly at competitive pressure - OpenAI’s GPT 5.6 Luna, a comparable lightweight model, is currently priced at $0.20 per million input tokens and $1.20 per million output tokens.

The release comes against a backdrop of delays at the higher end of Google’s model lineup. At I/O in May, Google said the flagship Gemini 3.5 Pro would ship in June; it has not appeared. Multiple Flash iterations have arrived in its place, with version numbers now edging toward 4.0. Reports cited by Ars Technica suggest Google’s coding capabilities have not kept pace with recent advances from rivals, and the company is also contending with an exodus of AI talent.

Availability is uneven at launch. Gemini 3.7 Flash is live in the Gemini API, AI Studio, and Gemini Enterprise. In the consumer Gemini app it powers the Gemini Spark agent, but only for AI Pro or Ultra subscribers; the standard chatbot interface continues to run on 3.6 Flash.

Practitioner takeaway: Developers and businesses using the Gemini API can switch to 3.7 Flash today and immediately halve their token costs under the introductory pricing, with real - if uneven - benchmark gains in coding and document tasks. The steep discount makes near-term experimentation low-risk. However, anyone waiting on the more capable 3.5 Pro for production use cases should plan around continued uncertainty - the model was promised two months ago and Google has given no revised timeline.

// read moreLLM-Gemini Plugin 0.33 Adds Gemini 3.7 Flash SupportAI ToolsNano Banana 2.1: Google Cuts Price in Half - Launch Day TestBenchmarksGoogle Releases AlphaGenome Atlas to Map Human Genome MutationsNews

Kinetyra