News
Updates, product notes and practical coverage for AI platforms and operators.
Google has released Gemini 4 Argon to AI Pro and Ultra subscribers, combining stronger reasoning with a more direct response style. The Gemini macOS app now embeds Google Antigravity for agentic coding, while a new task mode can continue work in the background and connect with more than 40 external apps.
Google Cloud has put its CLI behind a managed remote MCP server in public preview, letting Gemini Enterprise and other agents call gcloud and BigQuery commands without installing local binaries. The 30 September announcement emphasises caller-scoped IAM, isolated execution and audit logging. Teams should still govern which commands agents can run and verify preview limitations before production use.
Google’s 30 September Gemini Enterprise Agent Platform release notes mark the App Topology API as generally available. The update adds custom quick-query building and a simpler single-agent view in Agent Registry. For teams running multiple agents and tools, the practical gain is a clearer way to inspect dependencies, traffic and security context, provided the relevant services and permissions are configured.
Google has announced Gemini 4 Argon, a frontier model aimed at long-horizon coding, professional knowledge work and cyber defence. Its 30 September release begins with selected security defenders, not general public access. Google claims a one-million-token output limit and has disclosed introductory API pricing, while saying broader availability depends on feedback and stronger safeguards.
Google has introduced Ask in Earth Engine’s Code Editor, giving geospatial developers Gemini assistance for writing, explaining, debugging and optimising scripts. It reads the active script, imported assets and chat history and can show a code diff before changes are merged. Ask is available globally, but users must supply a Gemini API key and check any paid usage.
Google Cloud has detailed a managed reinforcement-learning fine-tuning service for Gemini. Developers provide prompts and a scoring function, while Google runs the training loop. The approach suits tasks whose outcomes can be checked more easily than ideal answers can be written. Teams should test reward quality, baseline performance, evaluation leakage and costs before adopting it.
Google has introduced Live Avatar for Gemini 3.8 Live in Gemini Enterprise, combining streaming video with speech and visual input. The company describes multilingual lip-sync, background tool calls and optional branded characters, with custom avatars restricted to allowlisted enterprises. The launch is separate from last week’s Live model release and needs careful identity and latency testing.
Google is rolling out new Connected Apps inside Gemini, spanning productivity tools such as Airtable and Linear, creative services including Adobe and Webflow, and lifestyle apps such as Peloton and SeatGeek. Users can connect services in settings or invoke them in chat. Access is rolling out, so permissions, account eligibility and action confirmations need checking.
Google has introduced Gemini 3.8 Flash TTS and Flash-Lite TTS for expressive speech generation. Flash focuses on custom voices and detailed performance direction; Flash-Lite targets high-volume use. Google describes consent checks for voice replication and SynthID watermarking. Developers should distinguish features rolling out now from Enterprise availability and voice remixing marked as coming soon.
Google has detailed Gemini features built into its new Googlebook laptop, including Magic Pointer, multilingual Rambler dictation and background task support. Pre-orders are open, with Australia among the listed launch markets for 5 October. These capabilities belong to the new device; existing Gemini users should not assume every feature is arriving on their current computer.
Google is rolling out spoken conversations, mobile audio capture and interactive learning overviews in Gemini Notebook. The update also describes new quiz formats and short video summaries, with features arriving on different schedules and plans. Students should check source citations and eligibility rather than assume every tool or promotion is already available to them.
Google Labs has expanded CC from a personal daily assistant into an experimental agent for households of up to six members. It can combine chosen information into shared briefs, calendars and tasks while each person controls what they share. Access is limited to adults in the United States; its connection to Gemini does not make it generally available in Australia.
Google has introduced Gemini 3.8 Live and Live Extended Thinking for real-time voice applications. The models support visual context, multilingual conversation and tool calls that can run while dialogue continues. The Live API release is distinct from rollout across consumer products; benchmark claims and regional access should be assessed against the actual use case.
Google has launched agentic video understanding for Gemini 3.7 Flash, 3.6 Flash and 3.5 Flash-Lite. The API feature lets models actively inspect video frames, audio and transcripts, aiming to cut token use and analysis cost while improving retrieval, counting and anomaly-detection tasks for developers working with long recordings.
Google Cloud has previewed Gemini Enterprise for Financial Services, combining purpose-built agent skills, secure data connectors and a Financial Research agent. The platform is designed for regulated workflows, but its useful deployment will depend on each institution’s data entitlements, governance and review processes, including human validation of consequential outputs.
Google Cloud has previewed Gemini Enterprise for Legal, a governed agent platform for law firms and legal departments. It combines legal skills, permission-aware connectors and partner agents for research, contract and compliance workflows, while keeping human legal judgment and existing data controls central across confidential, matter-specific work.
Google Cloud has introduced new Gemini Enterprise billing options and cost controls for agent workloads, including consumption pricing, pooled quotas, monthly caps and Flexible Savings Plans. The changes are designed to give organisations more predictable ways to scale AI agents while retaining centralised governance over spend.
Google has introduced Gemini 3.7 Flash, positioning it as a stronger workhorse model for coding and agents. The company reports gains across software engineering, web development and knowledge work, and has set an introductory price through year-end that is half the original Gemini 3.6 Flash price per million tokens.
Google Cloud’s Gemini Enterprise can now connect natively to ZoomInfo’s account, contact and buyer-intent data. The public-preview connector lets authorised agents query the GTM Context Graph in natural language, with access controls and audit information tied to customers’ existing Gemini Enterprise and ZoomInfo permissions.
Intel and Google Cloud have expanded their multi-year collaboration to deploy Gemini Enterprise across Intel and add agentic workflows to engineering and business operations. The agreement also extends Google Cloud infrastructure into silicon-development simulations, though financial terms, rollout dates and measured productivity gains remain undisclosed.