# Google DeepMind News

Read the latest articles and stories from DeepMind and find out more about our latest breakthroughs in cutting-edge AI research.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Introducing Gemini 4 Neon

DevFeed: [Introducing Gemini 4 Neon](<https://devfeed.tech/articles/introducing-gemini-4-neon-141911.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-gemini-4-neon/>)

Author: Koray Kavukcuoglu

Published: 2026-11-18T21:35:43Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [ai security](<https://devfeed.tech/topics/ai-security.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>), [google-antigravity](<https://devfeed.tech/topics/google-antigravity.md>)

Tags: [accounting](<https://devfeed.tech/tags/accounting.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agent-workflows](<https://devfeed.tech/tags/agent-workflows.md>), [agents-last-exam](<https://devfeed.tech/tags/agents-last-exam.md>), [antigravity](<https://devfeed.tech/tags/antigravity.md>), [api](<https://devfeed.tech/tags/api.md>), [artificial-analysis-intelligence-index](<https://devfeed.tech/tags/artificial-analysis-intelligence-index.md>), [automationbench](<https://devfeed.tech/tags/automationbench.md>), [engineering](<https://devfeed.tech/tags/engineering.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4-neon](<https://devfeed.tech/tags/gemini-4-neon.md>), [knowledge-work](<https://devfeed.tech/tags/knowledge-work.md>), [none](<https://devfeed.tech/tags/none.md>), [science](<https://devfeed.tech/tags/science.md>), [software-engineering](<https://devfeed.tech/tags/software-engineering.md>)

### AI overview

Gemini 4 Neon is positioned as a fast, cost-efficient model for complex coding, analytical, enterprise, and scientific workflows. The text reports strong benchmark results, including #1 on the Artificial Analysis Coding Agent Index and #2 on the Vals Index; long-running agent tasks, a 1-million-token context window, and four thinking levels let teams tune capability and token spend. It claims lower task costs than Carbon and competing frontier models, with results varying by setting and benchmark. Neon is offered through the Gemini API, Google AI Studio, Gemini for business, Antigravity, and the Gemini app, with access depending on subscription or program eligibility. The text also describes safeguards against misuse and prompt injection, plus internal evaluations of honesty and agent behavior; these are company-reported claims.

### Source excerpt

# Introducing Gemini 4 Neon Today, we're introducing Gemini 4 Neon -- bringing frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows. Neon ranks ahead of comparable models and #3 overall with a score of 61 on the Artificial Analysis Intelligence Index at its xhigh thinking level, landing just behind larger, slower models. On the Vals Index, Neon ranks #2 and lands on the Pareto frontier.

## Genie 4 is now available

DevFeed: [Genie 4 is now available](<https://devfeed.tech/articles/genie-4-is-now-available-141910.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/genie-4-is-now-available/>)

Author: Diego Rivas

Published: 2026-11-18T18:13:42Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [World models](<https://devfeed.tech/topics/world-models.md>), [generative-ai-tools](<https://devfeed.tech/topics/generative-ai-tools.md>), [synthid](<https://devfeed.tech/topics/synthid.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ambient](<https://devfeed.tech/tags/ambient.md>), [audio](<https://devfeed.tech/tags/audio.md>), [build](<https://devfeed.tech/tags/build.md>), [capabilities](<https://devfeed.tech/tags/capabilities.md>), [chaining](<https://devfeed.tech/tags/chaining.md>), [challenges](<https://devfeed.tech/tags/challenges.md>), [change](<https://devfeed.tech/tags/change.md>), [community](<https://devfeed.tech/tags/community.md>), [genie](<https://devfeed.tech/tags/genie.md>), [genie-4](<https://devfeed.tech/tags/genie-4.md>), [genie-4-world-model](<https://devfeed.tech/tags/genie-4-world-model.md>), [google](<https://devfeed.tech/tags/google.md>), [google-ai](<https://devfeed.tech/tags/google-ai.md>), [google-ai-ultra](<https://devfeed.tech/tags/google-ai-ultra.md>), [model](<https://devfeed.tech/tags/model.md>), [none](<https://devfeed.tech/tags/none.md>), [project-genie](<https://devfeed.tech/tags/project-genie.md>), [world](<https://devfeed.tech/tags/world.md>), [world-model](<https://devfeed.tech/tags/world-model.md>)

### AI overview

Project Genie is rolling out to eligible Google AI Ultra subscribers, with Google AI Pro members able to join a waitlist. Genie 4 enables real-time prompt-based changes to a world during exploration and adds adaptive native audio. Genie 3 remains available for longer sessions--up to three minutes per generation--world chaining, panorama-seeded spaces, and in-world prompt-driven events. A conversational Project Genie agent coordinates multimodal models to create initial visuals and background scores; Gemini can offer context-aware suggestions or trigger events during exploration. Users can explore curated and community worlds, remix them, and publish creations. Generated video and audio are SynthID-watermarked, and inputs and outputs undergo safety screening. The text says both models remain frontier research and identifies camera latency, intricate physical interactions, real-time audio precision, and long-term scene permanence as ongoing challenges.

### Source excerpt

We are rolling out a new Project Genie experience to eligible Google AI Ultra subscribers, opening access to our new Genie 4 world model.

## Introducing Gemini 4 Hydrogen

DevFeed: [Introducing Gemini 4 Hydrogen](<https://devfeed.tech/articles/introducing-gemini-4-hydrogen-130707.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-gemini-4-hydrogen/>)

Author: Koray Kavukcuoglu

Published: 2026-11-12T18:00:50Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [google-antigravity](<https://devfeed.tech/topics/google-antigravity.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [AI in frontend development](<https://devfeed.tech/topics/ai-in-frontend-development.md>), [design to code tools](<https://devfeed.tech/topics/design-to-code-tools.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-tasks](<https://devfeed.tech/tags/agentic-tasks.md>), [agentic-workflows](<https://devfeed.tech/tags/agentic-workflows.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-studio](<https://devfeed.tech/tags/ai-studio.md>), [airtable](<https://devfeed.tech/tags/airtable.md>), [android-studio](<https://devfeed.tech/tags/android-studio.md>), [antigravity](<https://devfeed.tech/tags/antigravity.md>), [api](<https://devfeed.tech/tags/api.md>), [app](<https://devfeed.tech/tags/app.md>), [argon](<https://devfeed.tech/tags/argon.md>), [business](<https://devfeed.tech/tags/business.md>), [carbon](<https://devfeed.tech/tags/carbon.md>), [daily](<https://devfeed.tech/tags/daily.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4](<https://devfeed.tech/tags/gemini-4.md>), [gemini-4-hydrogen](<https://devfeed.tech/tags/gemini-4-hydrogen.md>), [hydrogen](<https://devfeed.tech/tags/hydrogen.md>), [none](<https://devfeed.tech/tags/none.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

Gemini 4 Hydrogen is presented as a fast, efficient model for everyday coding and agentic work, available across Google developer, enterprise, consumer, and Search products. The article says it exceeds Argon on the Artificial Analysis Intelligence Index at high thinking (51 vs. 50), costs $1.50 per million input tokens and $7.50 per million output tokens across its 1M-token context window, and offers low, medium, high, and xhigh reasoning levels. In Google Antigravity, Orchestrator can switch between Carbon and Hydrogen; on Terminal-Bench 4.0 it reached near-Carbon success while using 24% less quota and finishing 19% faster on average. Against Gemini 3.8 Flash, Hydrogen is reported to finish tasks 2.5 times faster and score 57% vs. 18.2% on Terminal-Bench 4.1. The article also describes gains in coding, multimodal knowledge work, Search, and agent safety, while noting these performance and safety figures are company-reported.

### Source excerpt

Today, we're introducing Gemini 4 Hydrogen, our daily driver for agentic workflows

## Gemini 4 Carbon is now available

DevFeed: [Gemini 4 Carbon is now available](<https://devfeed.tech/articles/gemini-4-carbon-is-now-available-128943.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/gemini-4-carbon-is-now-available/>)

Author: Koray Kavukcuoglu

Published: 2026-11-10T16:03:47Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>), [AI in frontend development](<https://devfeed.tech/topics/ai-in-frontend-development.md>), [Data analysis](<https://devfeed.tech/topics/data-analysis.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [0-day](<https://devfeed.tech/tags/0-day.md>), [3d](<https://devfeed.tech/tags/3d.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-coding](<https://devfeed.tech/tags/agentic-coding.md>), [agentic-workflows](<https://devfeed.tech/tags/agentic-workflows.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-mode](<https://devfeed.tech/tags/ai-mode.md>), [ai-studio](<https://devfeed.tech/tags/ai-studio.md>), [alignment](<https://devfeed.tech/tags/alignment.md>), [alphafold](<https://devfeed.tech/tags/alphafold.md>), [alphafold-3](<https://devfeed.tech/tags/alphafold-3.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [android-studio](<https://devfeed.tech/tags/android-studio.md>), [antigravity](<https://devfeed.tech/tags/antigravity.md>), [carbon](<https://devfeed.tech/tags/carbon.md>), [developers](<https://devfeed.tech/tags/developers.md>), [enterprises](<https://devfeed.tech/tags/enterprises.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4-carbon](<https://devfeed.tech/tags/gemini-4-carbon.md>), [none](<https://devfeed.tech/tags/none.md>)

### AI overview

Gemini 4 Carbon is Google's higher-reasoning model for complex, long-running work, including multi-step coding, dense multimodal document analysis, scientific research, and cybersecurity. It uses more tokens than Gemini 4 Argon for deeper exploration and verification; Google positions Argon as the better fit for most software engineering and knowledge work. Carbon is available to paying developers and enterprises through the Gemini API and Gemini for business at Argon's introductory price, and is rolling out to Google AI Ultra subscribers in Antigravity, the Gemini app, and Search AI Mode. Google reports top benchmark results in several coding and automation evaluations, and #2 on its new multimodal knowledge benchmark at over six times lower average task cost than the top model. Its preview `/swarm` feature in Antigravity coordinates agents across isolated workspaces. Access for researchers and cyber defenders is subject to program eligibility and selection. Google reports stronger safeguards than Argon in several tests, while acknowledging that resistance to sophisticated prompt injection still needs improvement.

### Source excerpt

Today, we're making Gemini 4 Carbon available to developers, enterprises, scientists, and consumers.

## Introducing AlphaProof Code: a breakthrough in verifying complex software at depth

DevFeed: [Introducing AlphaProof Code: a breakthrough in verifying complex software at depth](<https://devfeed.tech/articles/introducing-alphaproof-code-a-breakthrough-in-verifying-complex-software-at-depth-128945.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-alphaproof-code/>)

Author: Pushmeet Kohli; Swarat Chaudhuri

Published: 2026-11-10T12:02:28Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Formal methods](<https://devfeed.tech/topics/formal-methods.md>), [Lean](<https://devfeed.tech/topics/lean.md>), [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [code security](<https://devfeed.tech/topics/code-security.md>), [Code](<https://devfeed.tech/topics/code.md>)

Tags: [after](<https://devfeed.tech/tags/after.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [all](<https://devfeed.tech/tags/all.md>), [also](<https://devfeed.tech/tags/also.md>), [apps](<https://devfeed.tech/tags/apps.md>), [automatically](<https://devfeed.tech/tags/automatically.md>), [banking](<https://devfeed.tech/tags/banking.md>), [breakthrough](<https://devfeed.tech/tags/breakthrough.md>), [bridging](<https://devfeed.tech/tags/bridging.md>), [code](<https://devfeed.tech/tags/code.md>), [complex](<https://devfeed.tech/tags/complex.md>), [core](<https://devfeed.tech/tags/core.md>), [lean](<https://devfeed.tech/tags/lean.md>), [proofs](<https://devfeed.tech/tags/proofs.md>), [research](<https://devfeed.tech/tags/research.md>), [signal](<https://devfeed.tech/tags/signal.md>), [software](<https://devfeed.tech/tags/software.md>), [verify](<https://devfeed.tech/tags/verify.md>)

### AI overview

Google DeepMind reports that AlphaProof Code formally verified key security properties of Signal's compiled cryptographic code, connecting protocol-level goals to the implementation running on devices. Human experts and a mix of models translated the goals into mathematical specifications, which experts reviewed; Gemini Carbon agents then generated Lean proofs, and agents audited them, flagging about one percent of the codebase for further human review. Lean machine-checked the proofs. The work took four weeks, and in tests with two previously fixed bugs reintroduced into an isolated code version, the system proved the buggy versions insecure. The account describes verification of specified key properties, not a guarantee that all possible flaws are absent.

### Source excerpt

AlphaProof Code agents wrote three million lines of proofs in Lean to verify Signal's core encryption.

## Opening the datasets behind our AI weather models

DevFeed: [Opening the datasets behind our AI weather models](<https://devfeed.tech/articles/opening-the-datasets-behind-our-ai-weather-models-133528.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/opening-the-datasets-behind-our-ai-weather-models/>)

Author: WeatherNext team

Published: 2026-11-10T11:13:04Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Dataset development](<https://devfeed.tech/topics/dataset-development.md>), [Climate / Weather / Ocean Modeling](<https://devfeed.tech/topics/climate-weather-ocean-modeling.md>), [geospatial](<https://devfeed.tech/topics/geospatial.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [applications](<https://devfeed.tech/tags/applications.md>), [climate](<https://devfeed.tech/tags/climate.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [developers](<https://devfeed.tech/tags/developers.md>), [earth](<https://devfeed.tech/tags/earth.md>), [global](<https://devfeed.tech/tags/global.md>), [imagery](<https://devfeed.tech/tags/imagery.md>), [live](<https://devfeed.tech/tags/live.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [models](<https://devfeed.tech/tags/models.md>), [satellite](<https://devfeed.tech/tags/satellite.md>), [satellites](<https://devfeed.tech/tags/satellites.md>), [science](<https://devfeed.tech/tags/science.md>), [train](<https://devfeed.tech/tags/train.md>), [water](<https://devfeed.tech/tags/water.md>), [weather](<https://devfeed.tech/tags/weather.md>)

### AI overview

SatWeave and PARDIG are being released with historical archives and live feeds for researchers and developers building weather, hydrology, and climate applications. SatWeave combines imagery from five geostationary satellites into a global mosaic that gives WeatherNext 3 a live view of clouds and storms. The satellites differ in map projections, file formats, sensors, and scan schedules of 10 to 30 minutes, making a coherent global map difficult; the excerpt says existing composites commonly omit incompatible details, use four or fewer channels, or update only every few hours. PARDIG is an AI-based global precipitation dataset used to train WeatherNext 3 precipitation forecasts.

### Source excerpt

Foundational precipitation and geostationary satellite datasets to observe the planet clearly.

## Gemini Robotics 2.5 accelerates real-world general dexterity

DevFeed: [Gemini Robotics 2.5 accelerates real-world general dexterity](<https://devfeed.tech/articles/gemini-robotics-2-5-accelerates-real-world-general-dexterity-128944.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/gemini-robotics-2-5-accelerates-real-world-physical-intelligence/>)

Author: Carolina Parada

Published: 2026-11-09T19:51:47Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Robot Manipulation](<https://devfeed.tech/topics/robot-manipulation.md>), [Robotics Simulation](<https://devfeed.tech/topics/robotics-simulation.md>), [Physical AI](<https://devfeed.tech/topics/physical-ai.md>)

Tags: [accelerated](<https://devfeed.tech/tags/accelerated.md>), [agi](<https://devfeed.tech/tags/agi.md>), [agile](<https://devfeed.tech/tags/agile.md>), [argon](<https://devfeed.tech/tags/argon.md>), [dexterity](<https://devfeed.tech/tags/dexterity.md>), [digital-twin](<https://devfeed.tech/tags/digital-twin.md>), [failures](<https://devfeed.tech/tags/failures.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4](<https://devfeed.tech/tags/gemini-4.md>), [gemini-4-argon](<https://devfeed.tech/tags/gemini-4-argon.md>), [gemini-robotics-2-5](<https://devfeed.tech/tags/gemini-robotics-2-5.md>), [google](<https://devfeed.tech/tags/google.md>), [model](<https://devfeed.tech/tags/model.md>), [models](<https://devfeed.tech/tags/models.md>), [real-world](<https://devfeed.tech/tags/real-world.md>), [reliability](<https://devfeed.tech/tags/reliability.md>), [robot](<https://devfeed.tech/tags/robot.md>), [robot-dexterity](<https://devfeed.tech/tags/robot-dexterity.md>), [robotics](<https://devfeed.tech/tags/robotics.md>), [safety](<https://devfeed.tech/tags/safety.md>), [vision-language-action](<https://devfeed.tech/tags/vision-language-action.md>)

### AI overview

Argon can use a real-world camera scene to generate varied simulated practice runs and build a digital twin. Combined with real training demonstrations, this is described as helping robots generalize more quickly to new scenes and objects. Researchers can also use video footage to hypothesize causes of task failures, then test improvements in simulation and on real robots. The excerpt provides no specific performance measurements or details about safety methods.

### Source excerpt

Our most capable vision-language-action model enhances robot dexterity, reliability, and safety

## Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips

DevFeed: [Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips](<https://devfeed.tech/articles/gemini-omni-1-2-flash-fast-and-cost-effective-with-longer-clips-121056.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/gemini-omni-12-flash-fast-and-cost-effective-with-longer-clips/>)

Author: Anish Nangia

Published: 2026-11-05T17:05:16Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [generative-ai-tools](<https://devfeed.tech/topics/generative-ai-tools.md>)

Tags: [1080p](<https://devfeed.tech/tags/1080p.md>), [4k](<https://devfeed.tech/tags/4k.md>), [ads](<https://devfeed.tech/tags/ads.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-studio](<https://devfeed.tech/tags/ai-studio.md>), [animation](<https://devfeed.tech/tags/animation.md>), [api](<https://devfeed.tech/tags/api.md>), [audio](<https://devfeed.tech/tags/audio.md>), [background](<https://devfeed.tech/tags/background.md>), [brand](<https://devfeed.tech/tags/brand.md>), [business](<https://devfeed.tech/tags/business.md>), [camera](<https://devfeed.tech/tags/camera.md>), [clip](<https://devfeed.tech/tags/clip.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [consistency](<https://devfeed.tech/tags/consistency.md>), [gemini-omni-1-2-flash](<https://devfeed.tech/tags/gemini-omni-1-2-flash.md>), [generation](<https://devfeed.tech/tags/generation.md>), [none](<https://devfeed.tech/tags/none.md>), [pricing](<https://devfeed.tech/tags/pricing.md>), [video](<https://devfeed.tech/tags/video.md>), [video-generation](<https://devfeed.tech/tags/video-generation.md>)

### AI overview

Gemini Omni 1.2 Flash adds native video generation and editing up to 15 seconds, with one-second duration increments, multi-turn edits, and stronger consistency when following image, motion-clip, and audio references. The text says pricing is 30% lower than Omni 1.1, at $0.07 per second for 720p; users can draft at 360p or 720p, edit details by conversation, then upscale to 1080p or 4K. It also describes improved speech, lip-sync, and continuous camera movement. The model is generally available through the Gemini API and Gemini for business, with rollout across Google products. The source notes ongoing work on fast sports action, rapid script changes, and complex physics; outputs carry a SynthID watermark, and Provisioned Throughput is available.

### Source excerpt

Gemini Omni 1.2 Flash brings native 15-second video generation and editing, stronger consistency, and 30% lower pricing.

## Gemini 4 Carbon: Our next frontier model

DevFeed: [Gemini 4 Carbon: Our next frontier model](<https://devfeed.tech/articles/gemini-4-carbon-our-next-frontier-model-106152.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/gemini-4-carbon-our-next-frontier-model/>)

Author: Koray Kavukcuoglu

Published: 2026-10-28T17:00:58Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [LLM evaluation / benchmarking](<https://devfeed.tech/topics/llm-evaluation-benchmarking.md>), [post-training](<https://devfeed.tech/topics/post-training.md>), [prompt injection](<https://devfeed.tech/topics/prompt-injection.md>), [Multi-step AI processes](<https://devfeed.tech/topics/multi-step-ai-processes.md>), [future of software](<https://devfeed.tech/topics/future-of-software.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-tasks](<https://devfeed.tech/tags/agentic-tasks.md>), [agentic-workflows](<https://devfeed.tech/tags/agentic-workflows.md>), [alignment](<https://devfeed.tech/tags/alignment.md>), [artifacts](<https://devfeed.tech/tags/artifacts.md>), [artificial-analysis](<https://devfeed.tech/tags/artificial-analysis.md>), [attack](<https://devfeed.tech/tags/attack.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [bugs](<https://devfeed.tech/tags/bugs.md>), [carbon](<https://devfeed.tech/tags/carbon.md>), [chain-of-thought](<https://devfeed.tech/tags/chain-of-thought.md>), [chain-of-thought-monitoring](<https://devfeed.tech/tags/chain-of-thought-monitoring.md>), [cheating](<https://devfeed.tech/tags/cheating.md>), [chrome](<https://devfeed.tech/tags/chrome.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [frontier-model](<https://devfeed.tech/tags/frontier-model.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4-carbon](<https://devfeed.tech/tags/gemini-4-carbon.md>), [model](<https://devfeed.tech/tags/model.md>), [none](<https://devfeed.tech/tags/none.md>)

### AI overview

Gemini 4 Carbon is presented as Google's strongest model, with reported gains in long-horizon engineering and cybersecurity. The article cites scores of 88.2% on SEC-bench Pro for vulnerability discovery, 54.8% on ExploitGym for exploitation, and 71.0% on CWE-bench for patching; internal testing reportedly found over 2x the secure-code generation success rate of Argon. It also describes safeguards and alignment work, including a 0.0% attack success rate on the evaluated Gray Swan indirect prompt injection benchmark and reduced unsupported claims, misalignment alerts, and overreach in internal tests. Rollout starts with a subset of cyber defenders, then trusted partners through the Fairwind Program; access for developers, businesses, and consumers is planned after further guardrail iteration. Introductory pricing matches Argon: $2 per million input tokens and $10 per million output tokens, with cached input tokens 95% cheaper. These are claims and results reported in the announcement.

### Source excerpt

Gemini 4 Carbon is our strongest model with state-of-the-art cybersecurity and long horizon capabilities.

## Google DeepMind and Singapore Government announce new AI tool to protect citizens against scams

DevFeed: [Google DeepMind and Singapore Government announce new AI tool to protect citizens against scams](<https://devfeed.tech/articles/google-deepmind-and-singapore-government-announce-new-ai-tool-to-protect-citizens-against-scams-86359.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/on-device-ai-scam-detection-mddi/>)

Author: JJ Geewax; Miguel de Andrés

Published: 2026-10-14T02:00:14Z

Content type: article

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Security, Privacy and Abuse Prevention](<https://devfeed.tech/topics/security-privacy-and-abuse-prevention.md>), [gemma](<https://devfeed.tech/topics/gemma.md>), [Responsibility & Safety](<https://devfeed.tech/topics/responsibility-safety.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [end-to-end-encryption](<https://devfeed.tech/tags/end-to-end-encryption.md>), [gemma](<https://devfeed.tech/tags/gemma.md>), [privacy](<https://devfeed.tech/tags/privacy.md>), [responsibility-safety](<https://devfeed.tech/tags/responsibility-safety.md>), [singapore](<https://devfeed.tech/tags/singapore.md>)

### AI overview

Google DeepMind and GovTech Singapore describe an on-device scam detection system for messaging platforms. It screens inbound messages with a lightweight classifier, then uses a fine-tuned Gemma model to evaluate conversation context. Training took place in a managed GCP environment using anonymized datasets, and the article says the system was tested against synthetic dialogues and local corpora. The approach is presented as a way to detect scams while keeping chat data on-device and preserving end-to-end encryption.

### Source excerpt

Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.

## Developing Gemini 4 Argon responsibly and safely

DevFeed: [Developing Gemini 4 Argon responsibly and safely](<https://devfeed.tech/articles/developing-gemini-4-argon-responsibly-and-safely-86356.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/developing-gemini-4-argon-responsibly-and-safely/>)

Author: Anca Dragan

Published: 2026-10-13T19:43:07Z

Content type: article

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [ai safety](<https://devfeed.tech/topics/ai-safety.md>), [Reinforcement learning](<https://devfeed.tech/topics/reinforcement-learning.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-control](<https://devfeed.tech/tags/agent-control.md>), [alter](<https://devfeed.tech/tags/alter.md>), [automated](<https://devfeed.tech/tags/automated.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [biology](<https://devfeed.tech/tags/biology.md>), [blocking](<https://devfeed.tech/tags/blocking.md>), [bug](<https://devfeed.tech/tags/bug.md>), [bugs](<https://devfeed.tech/tags/bugs.md>), [capabilities](<https://devfeed.tech/tags/capabilities.md>), [categories](<https://devfeed.tech/tags/categories.md>), [chain-of-thought](<https://devfeed.tech/tags/chain-of-thought.md>), [code](<https://devfeed.tech/tags/code.md>), [context](<https://devfeed.tech/tags/context.md>), [continuous-monitoring](<https://devfeed.tech/tags/continuous-monitoring.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [defense-in-depth](<https://devfeed.tech/tags/defense-in-depth.md>), [gemini-4-argon](<https://devfeed.tech/tags/gemini-4-argon.md>), [none](<https://devfeed.tech/tags/none.md>), [safety](<https://devfeed.tech/tags/safety.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

Google says Gemini 4 Argon's stronger cyber and biology capabilities prompted added safeguards. Its misuse defenses combine request screening, internal activation probes, and monitoring during responses, with intervention when needed; the company reports that external biological jailbreak tests unlocked under 1% of tested sub-tasks. For agent misalignment, a real-time monitor reviews actions and internal scratchpad, blocking severe boundary violations and flagging cases for human review. In training, the company avoids penalizing chain-of-thought: an experiment found that doing so reduced openly stated shortcuts while increasing silent shortcuts. Built-in adversarial training and security checks also reduced attack success to 0.7% across 15 adaptive attempts on Gray Swan's indirect prompt injection benchmark. The excerpt describes selected defenses and company-reported tests, not the full safety report.

### Source excerpt

Learn more about our approach to safety and security for Gemini 4 Argon.

## Gemini 4 Argon for Robotics

DevFeed: [Gemini 4 Argon for Robotics](<https://devfeed.tech/articles/gemini-4-argon-for-robotics-86357.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/gemini-4-argon-for-robotics/>)

Author: Kanishka Rao

Published: 2026-10-13T16:01:58Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Robotics](<https://devfeed.tech/topics/robotics.md>), [Robot Manipulation](<https://devfeed.tech/topics/robot-manipulation.md>), [Robotics Simulation](<https://devfeed.tech/topics/robotics-simulation.md>), [contact-rich manipulation](<https://devfeed.tech/topics/contact-rich-manipulation.md>), [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [Humanoid Robots](<https://devfeed.tech/topics/humanoid-robots.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [3D](<https://devfeed.tech/topics/3d.md>), [robot sense of touch](<https://devfeed.tech/topics/robot-sense-of-touch.md>), [World models](<https://devfeed.tech/topics/world-models.md>), [cURL](<https://devfeed.tech/topics/curl.md>)

Tags: [benchmark](<https://devfeed.tech/tags/benchmark.md>), [gemini-4-argon](<https://devfeed.tech/tags/gemini-4-argon.md>), [robotics](<https://devfeed.tech/tags/robotics.md>)

### AI overview

The source presents Gemini 4 Argon as a general-purpose physical reasoning model for robotics, citing an 87% score on the ERQA benchmark and describing workflows across control, simulation, annotation, and data generation. Examples include controlling a Spot robot and a simple arm, iteratively steering the Gemini Omni world model, tuning MuJoCo assets and contact behavior against recorded episodes, retargeting human motions, and producing task progress annotations. It also describes benchmark and simulation-generation prompts, including a claimed increase in real-world success from 56% to 75% and a reduction in completion time from 35 to 24 seconds. These are claims in the supplied material; it provides no independent validation. The closing text says developers can access Argon as `gemini-4-argon` through the Interactions API and `google-genai` SDKs.

### Source excerpt

Discover Gemini 4 Argon for robotics: leading the ERQA benchmark with physical reasoning, zero-shot control, Real2Sim, and agentic workflows.

## Gemini 4 Argon: Built to solve your hardest problems

DevFeed: [Gemini 4 Argon: Built to solve your hardest problems](<https://devfeed.tech/articles/gemini-4-argon-built-to-solve-your-hardest-problems-86358.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-gemini-4-argon/>)

Author: Koray Kavukcuoglu

Published: 2026-10-13T16:01:52Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [Multi-step AI processes](<https://devfeed.tech/topics/multi-step-ai-processes.md>), [Quantum Computing](<https://devfeed.tech/topics/quantum-computing.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [3d](<https://devfeed.tech/tags/3d.md>), [about](<https://devfeed.tech/tags/about.md>), [accelerate](<https://devfeed.tech/tags/accelerate.md>), [acceleration](<https://devfeed.tech/tags/acceleration.md>), [actions](<https://devfeed.tech/tags/actions.md>), [adaptive](<https://devfeed.tech/tags/adaptive.md>), [adk](<https://devfeed.tech/tags/adk.md>), [administrator](<https://devfeed.tech/tags/administrator.md>), [advanced](<https://devfeed.tech/tags/advanced.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [developers](<https://devfeed.tech/tags/developers.md>), [enterprises](<https://devfeed.tech/tags/enterprises.md>), [frontier](<https://devfeed.tech/tags/frontier.md>), [frontier-model](<https://devfeed.tech/tags/frontier-model.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4-argon](<https://devfeed.tech/tags/gemini-4-argon.md>), [model](<https://devfeed.tech/tags/model.md>), [none](<https://devfeed.tech/tags/none.md>), [problems](<https://devfeed.tech/tags/problems.md>), [solve](<https://devfeed.tech/tags/solve.md>)

### AI overview

Gemini 4 Argon is rolling out globally for developers, enterprises, and eligible Gemini subscribers, with API pricing of $2 per million input tokens and $10 per million output tokens; cached input tokens are 95% cheaper. The article presents it as a model for complex, multi-step work, citing examples including codebase refactoring, software performance improvements, vulnerability discovery, scientific research, and robotics. Google reports gains such as a 40% reduction in long-session preparation time in ADK and nearly doubled throughput for a quantum error-correction decoder. Access depends on user type and subscription, and the API supports a 1-million-token unified input-output limit with a new continuation protocol. Google says its safeguards include jailbreak defenses, real-time monitoring of agent actions, and training against prompt injection; the cited results are company-reported evaluations and examples.

### Source excerpt

We're giving developers, enterprises, scientists, and consumers access to Gemini 4 Argon, our latest frontier model.

## EmbeddingGemma 2: an open, lightweight multimodal embedding model

DevFeed: [EmbeddingGemma 2: an open, lightweight multimodal embedding model](<https://devfeed.tech/articles/embeddinggemma-2-an-open-lightweight-multimodal-embedding-model-65886.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/embeddinggemma-2-an-open-lightweight-multimodal-embedding-model/>)

Author: Sahil Dua

Published: 2026-10-06T19:57:04Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [gemma](<https://devfeed.tech/topics/gemma.md>), [Retrieval Augmented Generation (RAG)](<https://devfeed.tech/topics/retrieval-augmented-generation-rag.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [quantization](<https://devfeed.tech/topics/quantization.md>), [Transformers](<https://devfeed.tech/topics/transformers.md>), [sentence-transformers](<https://devfeed.tech/topics/sentence-transformers.md>), [llama.cpp](<https://devfeed.tech/topics/llama-cpp.md>), [Ollama](<https://devfeed.tech/topics/ollama.md>), [sglang](<https://devfeed.tech/topics/sglang.md>)

Tags: [8k](<https://devfeed.tech/tags/8k.md>), [apache-2-0](<https://devfeed.tech/tags/apache-2-0.md>), [browser](<https://devfeed.tech/tags/browser.md>), [code-search](<https://devfeed.tech/tags/code-search.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [cross-platform](<https://devfeed.tech/tags/cross-platform.md>), [data-privacy](<https://devfeed.tech/tags/data-privacy.md>), [databases](<https://devfeed.tech/tags/databases.md>), [embedding](<https://devfeed.tech/tags/embedding.md>), [embedding-model](<https://devfeed.tech/tags/embedding-model.md>), [embeddinggemma-2](<https://devfeed.tech/tags/embeddinggemma-2.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [lightweight](<https://devfeed.tech/tags/lightweight.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [none](<https://devfeed.tech/tags/none.md>), [open](<https://devfeed.tech/tags/open.md>)

### AI overview

EmbeddingGemma 2 is a 740-million-parameter, Apache 2.0 model that embeds text, code, images, video and audio in a shared space for local search and retrieval. It supports an 8K-token context, and its vectors can be shortened from 768 to 512, 256 or 128 dimensions, which the source says can reduce local vector storage and memory use by up to 6x. Optional encoders let text-only workloads use fewer parameters; on a Pixel 11 Pro, quantized text-only weights reportedly need about 191 MB of active RAM, versus about 567 MB for the full multimodal model. The source reports a 9.92-point improvement over EmbeddingGemma on MTEB Code and says local processing can support offline, privacy-first retrieval. Deployment options named include MediaPipe, LiteRT, transformers.js and WebGPU; model weights are available on Hugging Face and Kaggle.

### Source excerpt

Introducing EmbeddingGemma 2, an open multimodal embedding model optimized for privacy-first use cases

## Gemini 4 Argon: our next era of frontier intelligence

DevFeed: [Gemini 4 Argon: our next era of frontier intelligence](<https://devfeed.tech/articles/gemini-4-argon-our-next-era-of-frontier-intelligence-62875.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/gemini-4-argon-our-next-era-of-frontier-intelligence/>)

Author: Koray Kavukcuoglu

Published: 2026-09-30T20:01:45Z

Content type: news

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [future of software](<https://devfeed.tech/topics/future-of-software.md>)

Tags: [c](<https://devfeed.tech/tags/c.md>), [c-plus-plus](<https://devfeed.tech/tags/c-plus-plus.md>), [developers](<https://devfeed.tech/tags/developers.md>), [engineering-productivity](<https://devfeed.tech/tags/engineering-productivity.md>), [frontier-model](<https://devfeed.tech/tags/frontier-model.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-4-argon](<https://devfeed.tech/tags/gemini-4-argon.md>), [none](<https://devfeed.tech/tags/none.md>), [open-source-software](<https://devfeed.tech/tags/open-source-software.md>)

### AI overview

Google announces Gemini 4 Argon, a frontier model designed for complex, long-horizon workflows in software engineering, enterprise knowledge work, and cybersecurity defense. It is rolling out first to trusted cyber defenders, with broader access planned after further testing and guardrail work. Google describes internal uses including coding, memory optimization, and large codebase migrations, which it says will undergo auditing, emulation testing, and review before production.

### Source excerpt

Announcing Gemini 4 Argon, our frontier model for real-world coding, enterprise knowledge work, and cyber defense, rolling out soon.

## Introducing SynthID Bio

DevFeed: [Introducing SynthID Bio](<https://devfeed.tech/articles/introducing-synthid-bio-62876.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-synthid-bio/>)

Author: Pushmeet Kohli; David Stutz; Ali Cowen-Rivers; Jeremy Ratcliff

Published: 2026-09-30T15:03:07Z

Content type: article

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [synthid](<https://devfeed.tech/topics/synthid.md>), [watermarking](<https://devfeed.tech/topics/watermarking.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [proteins](<https://devfeed.tech/tags/proteins.md>), [science](<https://devfeed.tech/tags/science.md>), [synthid](<https://devfeed.tech/tags/synthid.md>), [watermarking](<https://devfeed.tech/tags/watermarking.md>)

### AI overview

SynthID Bio embeds watermarks in AI-generated biological designs to help track their provenance. The article describes it as a verification signal that could support DNA synthesis screening and help distinguish designs from trusted models with built-in safeguards.

### Source excerpt

Proof of concept for watermarking AI-generated proteins while preserving biological function.

## Introducing Gemini 3.8 Live with Live Avatar

DevFeed: [Introducing Gemini 3.8 Live with Live Avatar](<https://devfeed.tech/articles/introducing-gemini-3-8-live-with-live-avatar-59562.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-gemini-38-live-with-live-avatar/>)

Author: Shuo-yiin Chang

Published: 2026-09-24T16:20:39Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Conversational AI](<https://devfeed.tech/topics/conversational-ai.md>), [Real-time AI interactions](<https://devfeed.tech/topics/real-time-ai-interactions.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>)

Tags: [accessible](<https://devfeed.tech/tags/accessible.md>), [adapting](<https://devfeed.tech/tags/adapting.md>), [advanced](<https://devfeed.tech/tags/advanced.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-generated-content](<https://devfeed.tech/tags/ai-generated-content.md>), [ai-products](<https://devfeed.tech/tags/ai-products.md>), [api](<https://devfeed.tech/tags/api.md>), [api-documentation](<https://devfeed.tech/tags/api-documentation.md>), [asynchronous](<https://devfeed.tech/tags/asynchronous.md>), [audio](<https://devfeed.tech/tags/audio.md>), [avatar](<https://devfeed.tech/tags/avatar.md>), [background](<https://devfeed.tech/tags/background.md>), [brand](<https://devfeed.tech/tags/brand.md>), [capabilities](<https://devfeed.tech/tags/capabilities.md>), [conversational-ai](<https://devfeed.tech/tags/conversational-ai.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-3-8-live](<https://devfeed.tech/tags/gemini-3-8-live.md>), [live](<https://devfeed.tech/tags/live.md>), [none](<https://devfeed.tech/tags/none.md>), [real-time](<https://devfeed.tech/tags/real-time.md>)

### AI overview

Gemini 3.8 Live with Live Avatar pairs near real-time video generation and speech for enterprise agents that listen, see, and respond with synchronized expressions, lip movements, and fluid turn-taking. It can run tool calls and retrieve data in the background while dialogue continues. Its speech-to-speech synchronization supports transitions across 97 languages, with the text claiming no loss of video fidelity or visual drift. Gemini Enterprise includes preset avatars; custom avatars generated from reference images are available only through enterprise allowlisting. Generated audio and video carry an imperceptible SynthID watermark intended to help detect AI-generated content.

### Source excerpt

Introducing Gemini 3.8 Live with Live Avatar, which brings near real-time visual presence to Gemini's conversational AI.

## Advancing Private AI Compute with secure, server-side memory

DevFeed: [Advancing Private AI Compute with secure, server-side memory](<https://devfeed.tech/articles/advancing-private-ai-compute-with-secure-server-side-memory-58926.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/advancing-private-ai-compute-with-secure-server-side-memory/>)

Author: Google Private AI Compute Team

Published: 2026-09-23T16:00:57Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Cloud](<https://devfeed.tech/topics/cloud.md>), [trusted-execution-environment](<https://devfeed.tech/topics/trusted-execution-environment.md>), [cloud-infrastructure](<https://devfeed.tech/topics/cloud-infrastructure.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-systems](<https://devfeed.tech/tags/ai-systems.md>), [encrypted](<https://devfeed.tech/tags/encrypted.md>), [on-device](<https://devfeed.tech/tags/on-device.md>), [privacy](<https://devfeed.tech/tags/privacy.md>), [private-ai](<https://devfeed.tech/tags/private-ai.md>), [responsibility-safety](<https://devfeed.tech/tags/responsibility-safety.md>), [sealed](<https://devfeed.tech/tags/sealed.md>), [storage](<https://devfeed.tech/tags/storage.md>)

### AI overview

Google DeepMind describes a Private AI Compute update that adds persistent server-side memory for personal AI across devices. The architecture uses encrypted cloud storage, device-held cryptographic keys, and an isolated secure enclave that temporarily decrypts data to process requests before re-encrypting it.

### Source excerpt

Introducing private, server-side memory to Private AI Compute for personal AI.

## Gemini 3.8 text-to-speech says hello

DevFeed: [Gemini 3.8 text-to-speech says hello](<https://devfeed.tech/articles/gemini-3-8-text-to-speech-says-hello-58927.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/say-hello-to-gemini-38-text-to-speech/>)

Author: Leland Rechis

Published: 2026-09-23T15:25:14Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [voice ai](<https://devfeed.tech/topics/voice-ai.md>), [Speech Processing](<https://devfeed.tech/topics/speech-processing.md>), [Generative AI](<https://devfeed.tech/topics/generative-ai.md>), [generative-ai-tools](<https://devfeed.tech/topics/generative-ai-tools.md>), [synthid](<https://devfeed.tech/topics/synthid.md>), [Conversational AI](<https://devfeed.tech/topics/conversational-ai.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-studio](<https://devfeed.tech/tags/ai-studio.md>), [audio](<https://devfeed.tech/tags/audio.md>), [brand](<https://devfeed.tech/tags/brand.md>), [custom](<https://devfeed.tech/tags/custom.md>), [custom-voices](<https://devfeed.tech/tags/custom-voices.md>), [customization](<https://devfeed.tech/tags/customization.md>), [dubbing](<https://devfeed.tech/tags/dubbing.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-3-8-flash-lite-tts](<https://devfeed.tech/tags/gemini-3-8-flash-lite-tts.md>), [gemini-3-8-flash-tts](<https://devfeed.tech/tags/gemini-3-8-flash-tts.md>), [none](<https://devfeed.tech/tags/none.md>), [speech](<https://devfeed.tech/tags/speech.md>), [text-to-speech](<https://devfeed.tech/tags/text-to-speech.md>), [tts](<https://devfeed.tech/tags/tts.md>)

### AI overview

Google introduces Gemini 3.8 Flash TTS for creative voice design and line-by-line performance direction, and Gemini 3.8 Flash-Lite TTS for high-volume, cost-efficient dubbing, content creation, and voice agents. Both support expressive delivery controls; Flash TTS also offers voice creation from prompts, a library of 2,000+ voices, and replication from a 30-second sample with owner consent verification. Generated audio carries SynthID watermarks. Developers can try the models in Google AI Studio and use them through the Gemini API. Flash TTS is rolling out in Gemini Notebook; Flash-Lite in Google Vids. Enterprise API access is described as coming soon.

### Source excerpt

Gemini 3.8 Flash-Lite TTS and Gemini 3.8 Flash TTS are our most expressive audio models yet.

## Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking

DevFeed: [Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking](<https://devfeed.tech/articles/introducing-gemini-3-8-live-and-3-8-live-extended-thinking-26922.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-gemini-3-8-live-and-3-8-live-extended-thinking/>)

Author: Tom Ouyang

Published: 2026-09-15T17:05:57Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [Real-time AI interactions](<https://devfeed.tech/topics/real-time-ai-interactions.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-generated-content](<https://devfeed.tech/tags/ai-generated-content.md>), [ai-studio](<https://devfeed.tech/tags/ai-studio.md>), [api](<https://devfeed.tech/tags/api.md>), [artificial-analysis](<https://devfeed.tech/tags/artificial-analysis.md>), [audio](<https://devfeed.tech/tags/audio.md>), [big-bench-audio](<https://devfeed.tech/tags/big-bench-audio.md>), [build](<https://devfeed.tech/tags/build.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [enterprises](<https://devfeed.tech/tags/enterprises.md>), [extended-thinking](<https://devfeed.tech/tags/extended-thinking.md>), [fluid](<https://devfeed.tech/tags/fluid.md>), [for-developers](<https://devfeed.tech/tags/for-developers.md>), [frontier-models](<https://devfeed.tech/tags/frontier-models.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-3-8-live](<https://devfeed.tech/tags/gemini-3-8-live.md>), [live](<https://devfeed.tech/tags/live.md>), [none](<https://devfeed.tech/tags/none.md>)

### AI overview

Gemini 3.8 Live targets scalable, cost-efficient voice agents, with near-real-time visual input, transitions across 97 supported languages, and background tool or API calls that let conversation continue while work runs. Gemini 3.8 Live Extended Thinking is aimed at complex workflows: it reasons while speaking, using verbal acknowledgments and progress narration during multi-step tasks. The article reports Extended Thinking benchmark results, including a top Speech to Speech Quality Index score of 82.6, while describing Live as highly cost-effective and second in the Speech Agent Arena; these are company-reported results. Both models are rolling out through developer APIs and Google AI Studio, with enterprise access in private preview and consumer availability varying by product and subscription. Generated audio carries an imperceptible SynthID watermark.

### Source excerpt

Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking are our most advanced live dialogue models yet, built for natural conversation.

## Introducing WeatherNext 3, our most advanced and accurate global weather AI model

DevFeed: [Introducing WeatherNext 3, our most advanced and accurate global weather AI model](<https://devfeed.tech/articles/introducing-weathernext-3-our-most-advanced-and-accurate-global-weather-ai-model-6210.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-weathernext-3-our-most-advanced-and-accurate-global-weather-ai-model/>)

Author: The WeatherNext team

Published: 2026-09-03T15:02:08Z

Content type: article

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Climate / Weather / Ocean Modeling](<https://devfeed.tech/topics/climate-weather-ocean-modeling.md>), [Advanced AI workflows](<https://devfeed.tech/topics/advanced-ai-workflows.md>), [Google Maps](<https://devfeed.tech/topics/google-maps.md>), [BigQuery](<https://devfeed.tech/topics/bigquery.md>)

Tags: [3](<https://devfeed.tech/tags/3.md>), [advanced](<https://devfeed.tech/tags/advanced.md>), [africa](<https://devfeed.tech/tags/africa.md>), [agriculture](<https://devfeed.tech/tags/agriculture.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-model](<https://devfeed.tech/tags/ai-model.md>), [api](<https://devfeed.tech/tags/api.md>), [app](<https://devfeed.tech/tags/app.md>), [bigquery](<https://devfeed.tech/tags/bigquery.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [cloud-storage](<https://devfeed.tech/tags/cloud-storage.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [global](<https://devfeed.tech/tags/global.md>), [google](<https://devfeed.tech/tags/google.md>), [google-maps](<https://devfeed.tech/tags/google-maps.md>), [maps](<https://devfeed.tech/tags/maps.md>), [model](<https://devfeed.tech/tags/model.md>), [none](<https://devfeed.tech/tags/none.md>), [platform](<https://devfeed.tech/tags/platform.md>), [search](<https://devfeed.tech/tags/search.md>), [weather](<https://devfeed.tech/tags/weather.md>)

### AI overview

WeatherNext 3 uses hourly geostationary satellite mosaics and sparse weather-station observations to produce global forecasts at higher spatial resolution than WeatherNext 2: selected surface variables reach 5 km, and atmospheric variables such as wind reach 25 km. The article reports up to 60% CRPS improvement against IMERG, 30% against MRMS, and 10% against rain gauges at early lead times; it also says Google weather experiences will show up to 50% more accurate precipitation forecasts a day or more ahead. The model adds turbine-height wind and solar conditions for renewable-energy planning. Forecast data is offered through BigQuery, Earth Engine, and Cloud Storage, while the model is being integrated into Search, Gemini, Google Maps, and related services. The source cautions that weather remains unpredictable and directs users to local meteorological agencies for official forecasts and safety advisories.

### Source excerpt

WeatherNext 3, our most advanced global weather AI model, is now in Search, Gemini, Maps, Google Maps Platform, and Cloud.

## Proactive cyber defense for governments and enterprises

DevFeed: [Proactive cyber defense for governments and enterprises](<https://devfeed.tech/articles/proactive-cyber-defense-for-governments-and-enterprises-6231.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/proactive-cyber-defense-for-governments-and-enterprises/>)

Author: Four Flynn

Published: 2026-09-02T16:24:24Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [AI in cybersecurity](<https://devfeed.tech/topics/ai-in-cybersecurity.md>), [vulnerability management](<https://devfeed.tech/topics/vulnerability-management.md>), [cyber resiliency](<https://devfeed.tech/topics/cyber-resiliency.md>), [national security](<https://devfeed.tech/topics/national-security.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [advanced](<https://devfeed.tech/tags/advanced.md>), [agencies](<https://devfeed.tech/tags/agencies.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-threat](<https://devfeed.tech/tags/ai-threat.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [authentication](<https://devfeed.tech/tags/authentication.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [cyber](<https://devfeed.tech/tags/cyber.md>), [cyber-defense](<https://devfeed.tech/tags/cyber-defense.md>), [defense](<https://devfeed.tech/tags/defense.md>), [enterprises](<https://devfeed.tech/tags/enterprises.md>), [fairwind](<https://devfeed.tech/tags/fairwind.md>), [fairwind-program](<https://devfeed.tech/tags/fairwind-program.md>), [none](<https://devfeed.tech/tags/none.md>), [partners](<https://devfeed.tech/tags/partners.md>), [proactive](<https://devfeed.tech/tags/proactive.md>), [program](<https://devfeed.tech/tags/program.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

Google says its limited-access Fairwind Program gives selected Google Cloud customers, government agencies and cybersecurity partners access to Gemini 3.8 Flash Cyber with the CodeMender harness to find, verify and fix vulnerabilities. It says the combination can produce validated, deployment-ready patches in minutes within an organization's secure cloud, at lower operating cost than traditional frontier models. Initial access targets governments, critical infrastructure operators and core technology platforms; participants must restrict use to internal cybersecurity, incident response or penetration-testing staff and use protections such as multi-factor authentication. Google reports more than 650 participating partners. Other Google Cloud customers can use CodeMender with publicly available models hosted on Gemini Enterprise Agent Platform, alongside AI Threat Defense. The program is described as evolving with partner needs and expanding access over time.

### Source excerpt

The Fairwind Program is a limited access program for governments and trusted partners to use our cyber defense tools.

## Introducing Gemini 3.8 Flash and 3.8 Flash Cyber

DevFeed: [Introducing Gemini 3.8 Flash and 3.8 Flash Cyber](<https://devfeed.tech/articles/introducing-gemini-3-8-flash-and-3-8-flash-cyber-6199.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-gemini-3-8-flash-and-38-flash-cyber/>)

Author: Tulsee Doshi

Published: 2026-09-02T16:18:31Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [releases](<https://devfeed.tech/topics/releases.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [AI Bots](<https://devfeed.tech/topics/ai-bots.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [coding](<https://devfeed.tech/tags/coding.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [flash](<https://devfeed.tech/tags/flash.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [none](<https://devfeed.tech/tags/none.md>), [reasoning](<https://devfeed.tech/tags/reasoning.md>), [release](<https://devfeed.tech/tags/release.md>), [software-engineering](<https://devfeed.tech/tags/software-engineering.md>), [vulnerability](<https://devfeed.tech/tags/vulnerability.md>)

### AI overview

Google introduces Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, highlighting coding, multi-step reasoning, autonomous-agent use, cybersecurity capabilities, pricing, and benchmark claims.

### Source excerpt

Gemini 3.8 Flash and 3.8 Flash Cyber deliver next-generation intelligence for agentic workflows and cybersecurity.

## Introducing agentic video understanding with Gemini

DevFeed: [Introducing agentic video understanding with Gemini](<https://devfeed.tech/articles/introducing-agentic-video-understanding-with-gemini-6192.md>)

Original publisher: [Read original article](<https://deepmind.google/blog/introducing-agentic-video-in-gemini/>)

Author: Rohan Doshi

Published: 2026-09-01T17:08:51Z

Content type: release

Language: en

Sources: [Google DeepMind News](<https://devfeed.tech/sources/google-deepmind-news.md>)

Topics: [Google AI](<https://devfeed.tech/topics/google-ai.md>), [Google](<https://devfeed.tech/topics/google.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [ai](<https://devfeed.tech/tags/ai.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [api](<https://devfeed.tech/tags/api.md>), [audio](<https://devfeed.tech/tags/audio.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [cost](<https://devfeed.tech/tags/cost.md>), [developers](<https://devfeed.tech/tags/developers.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [feature](<https://devfeed.tech/tags/feature.md>), [flash](<https://devfeed.tech/tags/flash.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [google-ai](<https://devfeed.tech/tags/google-ai.md>), [models](<https://devfeed.tech/tags/models.md>), [none](<https://devfeed.tech/tags/none.md>), [performance](<https://devfeed.tech/tags/performance.md>), [reasoning](<https://devfeed.tech/tags/reasoning.md>), [retrieval](<https://devfeed.tech/tags/retrieval.md>), [tools](<https://devfeed.tech/tags/tools.md>), [transcripts](<https://devfeed.tech/tags/transcripts.md>), [video](<https://devfeed.tech/tags/video.md>)

### AI overview

Google launches agentic video understanding for Gemini Flash models. It uses native video tools to inspect relevant frames, audio, and transcripts, aiming to improve video-analysis accuracy while reducing token use and cost.

### Source excerpt

We're launching agentic video understanding across our latest Gemini models for improved accuracy and lower costs and token usage.

[Next page](<https://devfeed.tech/sources/google-deepmind-news.md?cursor=WyIyMDI2LTA5LTAxVDE3OjA4OjUxKzAwOjAwIiwgImQ5NmU5NDFmLTEyMGYtNDFjNC1iNzEwLWFkMzA1ZWUxNzVjNyJd>)