Google updates Gemini with faster Flash model
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Annelise Capossela/Axios
Google on Thursday debuted Gemini 3.7 Flash just three weeks after its last update to the workhorse version of its AI model.
Yes, but: The company has yet to deliver Gemini 3.5 Pro, a promised update to its larger version of Gemini as the search giant continues to trail Anthropic and OpenAI at the frontier.
Driving the news: Google said Gemini 3.7 Flash offers better performance in coding and knowledge work and is also more responsive to developers.
- "It better adapts to roadblocks, clarifies intent when needed, and follows instructions with greater fidelity," Google said in a blog post. "It thinks more diligently, putting in more effort into multi-step planning and tool calls."
- The company said its introductory pricing for the model will be 75 cents per million input tokens and $3.75 per million output tokens — half of the initial price for Gemini 3.6 Flash.
- The new model is also being added to Gemini Spark, the AI agent available to the Google AI Pro and Ultra subscribers.
Zoom in: The new model will also ship with updated guardrails to prevent misuse, Google says.
The big picture: Google's release comes amid a flurry of new models from U.S. competitors, including SpaceXAI's Grok, and Chinese providers, such as DeepSeek.
- While the company declined to discuss the fate of Gemini 3.5 Pro, it has said that it is training for Gemini 4 and excited by early results, so it's possible the company will scrap the release of 3.5 Pro and instead make the Gemini 4 Pro the next version of its large release.
