Google has launched a new AI model, Gemini 3.7 Flash, tailored to coding and automation tasks, Reuters reported on Aug. 13.

Google did not disclose when it will release its flagship Gemini 3.5 Pro model, the report said. Investors have been watching Gemini 3.5 Pro as a test of Google DeepMind’s AI competitiveness.

In July, Google said it was testing Gemini 3.5 Pro with partners and would release it soon.

Google said Gemini 3.7 Flash positions itself as a low-cost alternative for companies seeking to build autonomous AI systems that plan tasks on their own and use software tools to carry out multi-step work.

As a successor to Gemini 3.6 Flash, which was released 3 weeks ago, it improved performance on coding tasks such as debugging, resolving issues and generating production code.

Pricing was set at 75 cents per 1 million input tokens and $3.75 per 1 million output tokens. This introductory price, in effect through year-end, is about half the level of Gemini 3.6 Flash.

Reuters previously reported that Google co-founder Sergey Brin (세르게이 브린) has been urging key AI staff in recent months to focus on Gemini development. Last week, Google carried out a reshuffle of DeepMind’s leadership, with DeepMind CEO Demis Hassabis stepping down and vice president Koray Kavukcuoglu taking over. Around the same time, 2 co-leads of Gemini’s early technology left the company to co-found a separate startup.