Gemini App (gemini.google.com)
Consumer assistant for the webNot checked yetView source ↗
Live service health
Every Google AI surface that matters, verified from your browser. Each card probes its official endpoint, measures response time and reports exactly what it can and cannot confirm.
Checking…
Google AI Studio (aistudio.google.com)
Free prototyping workspace for GeminiNot checked yetView source ↗
Gemini API (generativelanguage.googleapis.com)
Developer endpoint for Gemini and ImagenNot checked yetView source ↗
Google Cloud Service Health (status.cloud.google.com)
Official cross-product incident dashboardNot checked yetView source ↗
Veo on Agent Platform
Cinematic video generation with native audioNot checked yetView source ↗
Imagen on Agent Platform
High-fidelity image generation with SynthIDNot checked yetView source ↗
What each service actually is
Six Google AI surfaces, six different jobs. Here is what each one does, how this monitor tests it, and where Google officially reports its incidents.
Google’s flagship consumer AI assistant, the chat home for Gemini models with Deep Research, Gems and Workspace ties
What it is: Google’s flagship consumer AI assistant, the chat home for Gemini models with Deep Research, Gems and Workspace ties.
How we check it: We verify that gemini.google.com answers your browser and time the response. Its availability depends on the Gemini web app and the accounts infrastructure behind your sign-in.
Where Google reports issues: Google reports Gemini Apps incidents through the Workspace status dashboard rather than a dedicated consumer feed, so this probe is the fastest early signal most users have.
The free browser workspace where developers prototype prompts, tune parameters and try Gemini, Veo and Imagen models before writing code
What it is: The free browser workspace where developers prototype prompts, tune parameters and try Gemini, Veo and Imagen models before writing code.
How we check it: We probe the app shell and also read Google’s own AI Studio status page at aistudio.google.com/status, which posts active issues for the Gemini API and Studio tooling.
Where Google reports issues: AI Studio publishes its own incident banner feed. When it reports issues, this dashboard treats your probe result as secondary and points you to the official banner.
the programmatic home of Gemini models and Imagen image generation for millions of apps
What it is: The Gemini API endpoint that serves generativelanguage requests — the programmatic home of Gemini models and Imagen image generation for millions of apps.
How we check it: The endpoint returns an error to unauthenticated requests on purpose. A fast answer of any kind proves the edge is alive; only silence or long stalls suggest trouble.
Where Google reports issues: AI Studio’s status page covers Gemini API incidents directly, and history includes real events such as the February 2026 global-endpoint error-rate incident recorded on Google Cloud Service Health.
Google’s official public dashboard for Google Cloud incidents, covering Agent Platform (formerly Vertex AI), the Gemini API in Cloud, and regional outages
What it is: Google’s official public dashboard for Google Cloud incidents, covering Agent Platform (formerly Vertex AI), the Gemini API in Cloud, and regional outages.
How we check it: We verify the dashboard is reachable from your browser. Its incident list is the authoritative record for platform-level Google AI outages.
Where Google reports issues: Known incidents recorded here include elevated Vertex AI Gemini API error rates on the global endpoint (27 February 2026). This monitor links you there rather than duplicating its data.
Veo is Google’s video generation model family, served on the Gemini Enterprise Agent Platform (formerly Vertex AI)
What it is: Veo is Google’s video generation model family, served on the Gemini Enterprise Agent Platform (formerly Vertex AI). Veo 3.1, released 2 April 2026, generates high-fidelity clips with native audio.
How we check it: We probe Veo’s official Agent Platform documentation and preview surface. Video generation is capacity-hungry, so Veo degradations usually appear as long queue times before they appear as errors.
Where Google reports issues: Agent Platform incidents surface on Google Cloud Service Health. If Veo probes look slow, the Cloud dashboard and your project quotas are the next places to check.
Imagen is Google’s text-to-image model family on the Agent Platform
What it is: Imagen is Google’s text-to-image model family on the Agent Platform. Imagen 4 brings photorealistic quality and best-in-class typography, with every output watermarked via SynthID.
How we check it: We probe Imagen’s official documentation hub on the Agent Platform. Image generation is lighter than video but still quota-bound, so slowdowns often trace to regional capacity.
Where Google reports issues: Older Imagen 3 and preview generations have documented deprecation timelines, so a failing call can be a lifecycle issue rather than an outage — the model versions page explains the difference.
The evolution of Google AI
From the Gemini announcement to the Agent Platform rename — the Google AI story in twelve steps.
10 May 2023Origin
Google flips to the Gemini era.
At Google I/O 2023 the company retires the Bard-first roadmap announcement and commits to Gemini, a natively multimodal model family built by Google DeepMind to power Search, Workspace and a new generation of developer tools.
Google blog ↗6 December 2023Gemini 1
Gemini 1 launches in three sizes.
Gemini 1.0 arrives in Ultra, Pro and Nano, powering Bard and Pixel devices, and Google rebrands its AI Studio around early Gemini API access for developers.
Google blog ↗February 2024Flagship
Gemini 1.5 Pro brings a million tokens.
The 1.5 generation introduces a breakthrough context window — one million tokens, later scaled further — making whole codebases, hours of video and document archives a single prompt.
Google blog ↗9 December 2024Agentic era
Gemini 2.0 opens the agentic age.
Gemini 2.0 launches with native image and audio output and tool use, introducing Project Astra prototypes, Jules the coding agent and deeper AI Overviews integration.
Google blog ↗March 2025Mainstream
Gemini 2.5 Pro takes the reasoning crown.
The 2.5 generation pairs thinking-first responses with top leaderboard scores, while the Gemini app surges up the charts and AI Studio becomes the default free playground for builders.
Google blog ↗6 February 2025Imagen in the API
Imagen 3 arrives for every developer.
Imagen 3 lands in the Gemini API and AI Studio, opening high-fidelity image generation to millions of developers and setting the template for text-to-image access that Imagen 4 later inherits.
Developers blog ↗18 November 2025Generation 3
Gemini 3 Pro begins the third era.
Google ships Gemini 3 Pro in preview across the Gemini app, AI Studio, Vertex AI and Search, marking its most capable multimodal reasoning generation to date.
Google blog ↗19 February 2026Refined flagship
Gemini 3.1 Pro sharpens reasoning.
Gemini 3.1 Pro refines the third generation’s reasoning stack, going from preview to wide availability and becoming the default behind Gemini app Deep Think experiences.
DeepMind model card ↗27 February 2026Real incident
A global-endpoint error spike.
Vertex AI Gemini API customers hit elevated error rates on the global endpoint. Google Cloud Service Health records the incident with start and end times — a live example of why the official dashboard matters.
Service Health incident ↗2 April 2026Veo 3.1
Veo 3.1 adds native audio on Agent Platform.
Veo 3.1 brings contextual, synchronized sound to video generation, documented and served through the Gemini Enterprise Agent Platform — the platform formerly known as Vertex AI.
Agent Platform docs ↗May 2026Platform shift
Vertex AI becomes the Agent Platform.
Google renames Vertex AI to the Gemini Enterprise Agent Platform, unifying agents, models and governance, with Imagen 4 and gemini-3-pro-image generations documented under the new home.
Agent Platform ↗September 2026Today
One AI stack, six critical surfaces.
From the consumer app to video and image generation APIs, Google AI now spans consumer, developer and enterprise layers — each with its own failure modes, and all watched by this dashboard.
Live monitor on this page ↗Live incident intelligence
Google publishes official feeds for its AI surfaces — which makes cross-checking possible. These verified references cover where Google AI service signals come from.
AI Studio status — Gemini API feed
Google’s own incident banner for the Gemini API and AI Studio lives at aistudio.google.com/status. It has carried real events, including Gemini API batch-endpoint issues. This is the first official source to check when API probes look wrong.
aistudio.google.com/status ↗Google Cloud Service Health
status.cloud.google.com is the authoritative dashboard for Agent Platform, Gemini API in Cloud and regional incidents, with a filterable product list, location views and a programmatic API for automated monitoring.
status.cloud.google.com ↗Workspace status dashboard
The Gemini web app has no dedicated consumer feed. Google reports Gemini Apps availability through the Google Workspace status dashboard, which is the closest official signal for consumer-facing outages.
Workspace dashboard ↗Model lifecycle, not always outages
Google publishes model versions and lifecycle dates for the Agent Platform. A failing Veo or Imagen call is sometimes a deprecated version rather than downtime — the model versions page tells you which.
Model versions page ↗Browser-side checks60-second refreshUTC timestampsTwo official feedsIndependent monitorAccessible motion
Built for clarity
Live means sourced, not simulated. This dashboard never converts an unreachable endpoint into a reassuring green state — when a check cannot complete, it says Unreachable and sends you to Google’s official feeds.
What a green card means
Your browser reached the official surface within the healthy latency window during the last check. It is a direct, recent observation — not a cached promise.
What a grey card means
The probe could not complete — often because a surface restricts cross-origin browser requests. That is a statement about this dashboard's visibility, not about Google's health, so no verdict is issued.