Gemini 3.7 Flash 发布：编码、知识工作与网页开发能力提升

The coding improvements include “strong gains” over 3.6 Flash for debugging and issue resolution. Specifically, the DeepSWE v1.1 benchmark goes from 49.0% to 65.3%, while it’s 34.4% to 43.6% on FrontierCode 1.1 Main. For web development, Gemini 3.7 Flash “generates more functional layouts and feature-complete apps in fewer prompts.” Last month’s model had an Elo score of 1538 on Arena.ai’s WebDev Arena, and this release comes in at 1588. Advertisement - scroll for more content > For UI generation, the model shows high design adherence and parity based on a reference input, whether it’s a screenshot, an image, or a full design system. In finance, law, biosciences, and other “knowledge-dense fields,” there are improvements in reasoning and accuracy: [...] Go to the 9to5Google home page Apps & Updates Gemini Gemini app # Gemini 3.7 Flash launches three weeks after last model, live in Spark Avatar for Abner Li Abner Li | Aug 13 2026 - 10:00 am PT 7 Comments Just three weeks

This release comes just three weeks after Gemini 3.6 Flash, and is a direct result of developer feedback and algorithmic innovations that we look forward to bringing to future models. 3.7 Flash delivers substantial improvements across software engineering, knowledge work, and web development workflows — with an introductory price of half the original 3.6 Flash cost per million tokens. ## Better intelligence for complex workflows 3.7 Flash shows strong gains over 3.6 Flash in coding tasks like debugging and issue resolution. It also achieves higher first-pass code accuracy and has improved performance in generating production-ready code as seen in FrontierCode 1.1 Main (43.6% vs 34.4%) and DeepSWE v1.1 (65.3% vs 49.0%). [...] In web development, 3.7 Flash generates more functional layouts and feature-complete apps in fewer prompts. For UI generation, the model shows high design adherence and parity based on a reference input, whether it’s a screenshot, an image, or a full design syste

The release comes only three weeks after Gemini 3.6 Flash and incorporates developer feedback alongside new algorithmic improvements. Google said Gemini 3.7 Flash provides substantial gains across software engineering, document-heavy knowledge work, and web development while launching at an introductory price equal to half the original Gemini 3.6 Flash cost per million tokens. Coding is one of the largest areas of improvement. On FrontierCode 1.1 Main, Gemini 3.7 Flash scored 43.6%, compared with 34.4% for Gemini 3.6 Flash. On DeepSWE v1.1, the new model scored 65.3%, up from 49.0%. Google said the model is better at debugging, issue resolution, first-pass code accuracy, and generating production-ready software. [...] × Search Pulse 2.0 # Google Launches Gemini 3.7 Flash With Lower Pricing And Stronger Coding And Agent Capabilities By Amit Chowdhry Pulse 2.0 Google Launches Gemini 3.7 Flash With Lower Pricing And Stronger Coding And Agent Capabilities Follow Pulse 2.0 on LinkedI

Skip to content Techgenyz More + # Google Gemini 3.7 Flash Takes a Big Leap Forward With Better Coding and Automation By Anjali Mishra Google Gemini 3.7 Flash Representational image based on an official image | Techgenyz ## Highlights Google launched Gemini 3.7 Flash just three weeks after releasing Gemini 3.6 Flash. The new model focuses on coding, AI agents, knowledge work, and web development. Google says Gemini 3.7 Flash delivers significant performance gains across software engineering and automation tasks. Internal benchmarks show improvements in FrontierCode, DeepSWE, WebDev Arena, GDP.pdf, and AutomationBench. [...] Google has made an official announcement regarding its newly launched Gemini 3.7 Flash. This is another update to Google’s growing range of Artificial Intelligence models. The latest product was launched merely three weeks after the company announced Gemini 3.6 Flash, highlighting Google’s increasingly aggressive development cycle for its lightweight AI

Gemini 3.7 Flash looks strong for coding on Google's own benchmarks (Google). It scored 43.6% on FrontierCode 1.1 and 65.3% on DeepSWE v1.1 (Google). These are vendor-run numbers, so run a short pilot on your own code before you switch. Gemini 3.7 Flash beats Gemini 3.6 Flash on every benchmark Google published (Google). That includes coding, web development, and agent tasks. It is also cheaper during the intro window through December 31, 2026 (Google). No. All five benchmark scores come from Google's own charts (Google). No independent, same-generation head-to-head has been published, so treat the numbers as vendor-run and pilot on your own tasks. [...] A large window helps with big codebases and long documents. It lets the model hold more of the problem at once, which supports the coding and agent focus. Knowledge work is a supporting strength. The GDP.pdf gain points to better long-document handling (Google). That helps with research, summaries, and analysis over big files. Cod

past couple of months and making those models the default inside of Chat GPT. By default, the model that's used is their best model. And Google has really fallen behind here, and the optics of this aren't great. Now, of course, they are Google, and I have full confidence they have all the resources they need to pull back ahead and to catch up to the competition. We just have to wait for that to happen. In the meantime, though, 3.7 Flash, like I said, is a great model. And they've said that it's stronger for coding, knowledge work, and web development. And looking at the benchmarks, you can see that there are nice improvements to it. So, you're going to notice a difference when you're switching from 3.6 Flash to 3.7 Flash. And the last update from Google is to Pexels. So, Pexels is a tool