Gemini 3.7 Flash: Google Halves the Price of Its Newest Model Until Year-End
Trending Topics Jakob Steinschaden ● Covered by 7 sources
Google just launched Gemini 3.7 Flash and cut its price in half until year-end. It’s aimed at coding and agents, but the real story is the cheap API window.
Based on reporting by Trending Topics, Jakob Steinschaden — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Google released Gemini 3.7 Flash on August 13, just three weeks after Gemini 3.6 Flash. The new model sits in the Flash line and is aimed at coding, agentic workflows, and document processing inside companies. The headline is not the benchmark sheet. It is the price tag.
Until December 31, 2026, the API costs are set at $0.75 per million input tokens and $3.75 per million output tokens. On January 1, 2027, they rise to $1.50 and $7.50. Google says the same discount also applies retroactively to Gemini 3.6 Flash, which only arrived at the higher prices in late July. For teams running agents in production, that matters more than a single scorecard, because token bills pile up quickly.
Google describes 3.7 Flash as its most intelligent workhorse model yet for coding and agents. The company says it handles roadblocks better, asks for clarification more often when tasks are unclear, and follows instructions more closely. For planning and tool use, it is supposed to spend more compute. It also pushes through at 340 tokens per second, which is the kind of speed real-time systems want.
The model card says this is not a fresh pretraining run. It is a refinement of 3.6 Flash with algorithmic changes to the reasoning core. The context window is one million tokens, maximum output is 64,000 tokens, and the knowledge cutoff stays at March 2026. Google’s own numbers show gains over 3.6 Flash on FrontierCode, DeepSWE v1.1, WebDev Arena, GDP.pdf, and AutomationBench, but those figures have not been independently checked.
And the table still has holes. On Terminal-bench 2.1, Gemini 3.7 Flash trails GPT-5.6 Terra, and OpenAI also leads on Terminal-bench 3.0 and OSWorld-2.0. Claude Sonnet 5 beats it on the multimodal Agent’s Last Exam. Google does look stronger on AutomationBench and GDP.pdf, but the bigger picture is that 3.7 Flash is still not the top model across the board. The flagship Gemini 3.5 Pro still has no release date, which is becoming a familiar Google habit: flashy flashes, missing crown jewels.
My take — AI-written commentary, not fact-checked reporting
This is classic Google: squeeze the price, ship the small model, and leave everyone staring at the missing flagship. Cheap tokens are nice, but the real strategy is to lock teams into production before they notice the ladder above Flash is still foggy. Open model fans will have fun pointing out that the “best” Google offer still isn’t best everywhere; they’re not wrong.
Read more about this at: Trending Topics
