Google shipped a cheaper Flash model on Wednesday. The Pro model it promised in May still has not arrived.
Gemini 3.7 Flash landed on 13 August, three weeks after 3.6 Flash. The coding gains are large: DeepSWE v1.1 rose from 49.0% to 65.3%, FrontierCode 1.1 from 34.4% to 43.6%, and on AutomationBench it scored 30.4% against GPT-5.6 Terra's 23.6% and Claude Sonnet 5's 10.7%. Until the end of 2026 it costs $0.75 per million input tokens and $3.75 per million output — half what the model it replaces cost at launch. It is live in the API, AI Studio, Android Studio, Antigravity and Gemini Enterprise, and in the Gemini app for Pro and Ultra subscribers. What Google still has not shipped is Gemini 3.5 Pro, the flagship Sundar Pichai said in May would arrive 'next month'; it has now missed three targets.
Why it mattersRead the two halves together. Google is shipping quickly and cutting prices at the cheap end of its lineup while the expensive end stays empty — the shape you get when the big model is in trouble and the small ones are fine. On Tuesday we reported an analyst claim that 3.5 Pro had been quietly cancelled; this launch neither confirms nor refutes it, but it does not make the claim look worse. For anyone actually choosing a model this month the practical news is simpler: a workhorse that codes far better than its predecessor, at half its predecessor's launch price.
Read in the app — free, in 9 languages
Related stories
Reflection will hand out a 501-billion-parameter model for free.2026-10-06Microsoft's new transcriber starts writing before you finish speaking.
2026-10-05An American lab is about to give away a China-class open model.
2026-10-05A video site now gives away the best-scoring open translation model.
2026-10-05OpenAI will ship a Codex upgrade daily for 28 days or reset limits.
2026-10-05