Gemini Spark review, pricing and limits

Gemini Spark is best for Google Workspace power users who want an assistant that keeps triaging email and calendar work after they close the laptop, replacing manual inbox checking with drafts held for approval. It launched with just three MCP connectors, Canva, OpenTable, and Instacart, so it fits Workspace-centric routines more than broad cross-app automation today.

Launched May 19, 2026 at Google I/O, Gemini Spark is Google's persistent AI agent, running on dedicated Google Cloud virtual machines instead of stopping when a chat tab closes. It manages Gmail, Calendar, Docs, and Drive, drafting actions like email replies for human approval rather than sending them autonomously.

Maker: Google · Autonomy: semi autonomous · Maturity: BETA

Underlying models: Gemini 3.5 Flash

About Gemini Spark

Gemini Spark is a persistent AI agent from Google, announced May 19, 2026 at Google I/O alongside a new Gemini model. Unlike the standard Gemini chat assistant, which stops the moment a tab closes, Spark runs continuously on Google's own cloud servers, sometimes for days at a stretch, so it can keep working on email, scheduling, and document tasks without being reopened. It sits inside the Gemini app as the proactive counterpart to Google's reactive chat assistant, part of a shift Google describes as moving Gemini from a tool that surfaces data to an agent that acts on it. That persistence is paired with Antigravity 2.0, Google's internal agent runtime that adds goal persistence, task decomposition, tool orchestration, and state recovery on top of the model. Spark can hold a goal like keeping an inbox triaged across days, decide when to act, call tools like Gmail and Calendar, and recover if interrupted. Consequential actions, like sending an email, are held as drafts for a person to approve, which makes it semi-autonomous rather than fully autonomous. Spark works best for people already living inside Google Workspace who face high email and calendar volume, such as small business owners and executive assistants. It monitors Gmail's AI Inbox, drafts personalized replies using context from past emails, Docs, and the calendar, and creates calendar events directly from email content. At launch it also connects over the Model Context Protocol to a small set of outside services for design, dining, and grocery tasks. Enterprise IT teams that need narrowly scoped, audited permissions are a weaker fit: security researchers have flagged Spark's broad, always-on Workspace access as carrying execution risk beyond normal data exposure. Spark isn't sold on its own: it's bundled into Google AI Ultra, and Google used the launch to cut Ultra's price to give the new agent a mass-market entry point. As of launch, Spark is available only to US-based, English-language Ultra subscribers, rolling out first to trusted testers and then more broadly. There's no free tier and no public developer API for Spark itself. Google has said more MCP connectors are coming over the summer of 2026, and broader rollout to AI Pro subscribers and Workspace business previews is planned for later in the year. International availability will also be shaped by the EU AI Act's consumer-agent obligations, which take effect August 2, 2026, so Spark is likely to stay a US-only, Workspace-centric agent for some time yet.

Pricing

Not sold separately. Bundled into Google AI Ultra: $99.99/mo entry tier (cut 60% from $249.99 at Spark's launch, ~5x AI Pro usage limits, 20TB storage, YouTube Premium) or AI Ultra Premium at $200/mo (down from $250, ~20x AI Pro limits). No free tier and no standalone Spark pricing.

Key Features

• 24/7 Persistent Runtime: Spark runs on dedicated Google Cloud virtual machines with a lifecycle measured in hours or days, continuing tasks even when your phone is locked or laptop is shut, unlike session-based chat agents.
• Gmail AI Inbox: Spark monitors incoming email and writes personalized draft replies using context from your Gmail, Docs, and Calendar, holding each draft for your approval before anything sends.
• Daily Brief: A continuously updated digest pulled from Gmail, Calendar, and Tasks surfaces what needs your attention each morning without you having to ask.
• Cross-App Workspace Reasoning: Spark cross-references Gmail, Calendar, Docs, Sheets, Slides, Drive, and Maps, for example creating calendar events directly from email content.
• MCP Third-Party Connectors: At launch Spark connects to Canva, OpenTable, and Instacart over the Model Context Protocol, with GitHub, Notion, Slack, Adobe, Samsung, Spotify, and CapCut planned for summer 2026.
• Built on Gemini 3.5 Flash: Spark's reasoning runs on Gemini 3.5 Flash at roughly 284.2 tokens per second, about 4x faster than comparable frontier models, wrapped in the Antigravity 2.0 agent runtime.

Strengths

• Its persistent runtime is a genuine point of difference from most AI Workspace assistants, which only work while a tab or app stays open.
• Access got meaningfully cheaper at launch: Google slashed the AI Ultra price to make room for Spark, so the always-on agent no longer requires the plan's priciest historical tier.
• Native, no-setup access to the full Google Workspace stack (Gmail, Calendar, Docs, Sheets, Slides, Drive, Maps), so it can act on real account data from day one.

Weaknesses

• Still a US-based, English-only beta, with no international rollout date announced.
• Requires a paid Google AI Ultra subscription; there's no free tier, no standalone price for Spark alone, and no public API.
• Spark's always-on Workspace access raises permission risk: security researchers note it can inherit broad folder and group access in Gmail, Docs, and Drive without evaluating business context.

Visit Gemini Spark Official Site