Google pushes Gemini into workflows
Google released a native Gemini app for macOS, added prepay billing for the Gemini API and is rolling out agentic features that let Gemini act across email, files and tasks. The updates also include Gemini 3.1 Flash TTS for expressive speech and early tests of 'Agent Mode' inside Workspace that can use local files and connected services to complete broader goals. These moves package desktop presence, easier commercial billing and agentic capabilities into a single push toward workflow integration. (workspaceupdates.googleblog.com) (blog.google) (blog.google) (androidauthority.com)
Google is moving Gemini from a browser tab into day-to-day work, starting with a native Mac app and new tools that let the model take actions across tasks. (workspaceupdates.googleblog.com) Google said on April 15 that the macOS app lives in the menu bar and opens with the Option-Space keyboard shortcut, putting Gemini “right where you work” on Apple desktops. The company said the app is available to Google Artificial Intelligence Pro, Google Artificial Intelligence Ultra, and qualifying Google Workspace business and education users. (workspaceupdates.googleblog.com) On the same day, Google added prepay billing for the Gemini application programming interface in Google Artificial Intelligence Studio. The company said new United States Google Cloud billing accounts can buy credits up front now, with a global rollout planned “in the coming weeks.” (blog.google) Google also introduced Gemini 3.1 Flash text-to-speech on April 15, a model for turning text into spoken audio with tone controls built into prompts. Google said it supports more than 70 languages and adds SynthID watermarking to mark audio as artificial intelligence-generated. (blog.google) An artificial intelligence assistant usually answers one prompt at a time; an “agent” is meant to carry out a broader goal by breaking it into steps, using tools, and returning with a finished result. A test version spotted this week inside Gemini Enterprise adds tabs for Goals, Agents, Connected apps, Files, and a Tasks dashboard, pointing to that fuller model of work. (androidauthority.com) That push has been building inside Google Workspace for months. In February, Google added Google Chat as a data source for the Gemini app alongside Gmail and Google Drive, and in late February it said conversation history was coming to the Gemini side panel in Workspace apps. (workspaceupdates.googleblog.com 1) (workspaceupdates.googleblog.com 2) (workspaceupdates.googleblog.com 3) Google has been making the same case to developers: Gemini should be easier to buy, cheaper to control, and better at multi-step work. Last month, Google added project spend caps, revised usage tiers, and new cost dashboards for the Gemini application programming interface before this week’s prepay option. (blog.google 1) (blog.google 2) The company has also been expanding agent features beyond Workspace. Google said yesterday that Gemini Command Line Interface now supports parallel “subagents,” separate specialist helpers that can work on multiple tasks at once under one main session. (developers.googleblog.com) Google has not publicly announced a full Workspace-wide launch date for the leaked Agent tab. But the April 15 releases show the direction clearly: desktop access, prepaid developer billing, voice output, and agent-style task handling are arriving as one package around Gemini. (workspaceupdates.googleblog.com) (blog.google 1) (blog.google 2) (androidauthority.com)
