ModelsThe Decoder· Jun 25, 2026
Google Adds Computer Control Feature to Gemini 3.5 Flash
Google has built computer operation capabilities directly into its Gemini 3.5 Flash model. The update allows the model to autonomously manage computers, browsers, and mobile devices through screen interaction. It reaches 78.4 on the OSWorld benchmark, matching GPT-5.5 results. The feature is accessible via the Gemini API for tasks such as software testing and automation.
Key points
- →Computer Use integrated into Gemini 3.5 Flash for direct device control
- →Model can operate computers, browsers, and mobile devices independently
- →Scores 78.4 on OSWorld benchmark, comparable to GPT-5.5
- →Gemini API enables developer use for testing and office automation
Mentioned
GoogleGemini 3.5 FlashGPT-5.5Gemini API
Related stories
Google Losing Top AI Researchers to Rival CompaniesThe Decoder · Research→AI Researchers Depart Google for Anthropic and RivalsTechCrunch · Research→OpenAI Updates GPT-5.5 Instant for Better Intent RecognitionThe Decoder · Models→Snowflake CEO Reports GLM-5.2 Matches Opus 4.7 at Lower CostThe Decoder · Models→Research Shows Reasoning Improves Knowledge Recall in Large Language ModelsGoogle Research · Research→Google Home Speaker Shows Strong Voice Detection in TestsThe Verge · Hardware→
This is an original summary by Dhanasvi's agents based on The Decoder's public feed. For the complete article, visit the original source. Trademarks and article copyright belong to their owners.