Original briefings. Zero spin.
Every story is an original briefing written from 60+ sources across the spectrum — sources linked so you can verify it yourself.
tech
Low Interest (4/10)
May 19, 2026 at 07:13 PM
Google Cuts AI Ultra Price in Half, Launches Spark Agent and Omni Video Model After I/O Keynote
Google Cuts AI Ultra Price in Half
Google cut its AI Ultra plan from $250 to $200 per month, according to Engadget. It also introduced a new $100-per-month Ultra tier aimed at developers and creative professionals — half the old entry price for the top subscription. The previous $250 price point wasn't sticking.
The $100 tier gets you 5x the usage limits of AI Pro, priority Antigravity access, 20TB of cloud storage, and Gemini Spark when it launches. The $200 tier gets 20x limits and Project Genie access. The old $250 plan is discontinued.
For context, Anthropic and OpenAI both run tiered plans at the $100 and $200 levels, according to TechCrunch. Google matched that structure. It's competitive pricing, not a new approach.
Gemini Spark: Autonomous Agent for Always-On Tasks
Gemini Spark is a 24/7 personal AI agent running on Gemini 3.5 Flash, hosted on Google Cloud virtual machines. According to VentureBeat, it accepts instructions — "email my boss a status update pulling figures from our shared spreadsheet" — and executes them across Gmail, Docs, Sheets, and Slides without further input. Your laptop can be closed. Your phone locked. Spark keeps working.
Google Labs VP Josh Woodward told TechCrunch it feels like "tossing things over your shoulder" — Spark catches them and gets the job done.
Trusted testers get it this week. Google AI Ultra subscribers in the US get beta access next week, per CNBC. Third-party integrations — Canva, OpenTable, Instacart — are live now via Model Context Protocol (MCP). Browser control through Chrome and the ability to text or email Spark directly are coming "in the coming weeks," according to Engadget.
Wired noted Google's structural advantage here: it already has your email. Spark's Gmail integration is native, not bolted on. That's a moat OpenAI and Anthropic's Claude Cowork don't have.
The key concern — which most coverage understated — is that you're giving an autonomous agent access to your entire digital life. Google says Spark will ask permission before "high-stakes actions like spending money or sending emails." That's a guardrail, not a guarantee.
Gemini 3.5 Flash: Speed and Performance Claims
Google DeepMind CTO Koray Kavukcuoglu told TechCrunch that Gemini 3.5 Flash outperforms Gemini 3.1 Pro — their flagship from five months ago — on nearly every benchmark. Specific numbers per VentureBeat: 76.2% on Terminal-Bench 2.1, 83.6% on MCP Atlas, 84.2% on CharXiv Reasoning.
Speed claim: four times faster than comparable frontier models. An optimized version runs 12 times faster at equivalent quality, per TechCrunch.
CEO Sundar Pichai told reporters that companies running roughly one trillion tokens per day on Google Cloud could save more than $1 billion annually by shifting 80% of workloads to Flash. Google provided its own internal calculations. Independent enterprise audits will determine if those numbers hold.
Gemini 3.5 Pro, the heavier reasoning model, is being used internally but won't ship publicly until next month, per CNBC.
Gemini Omni and Video Generation
Gemini Omni Flash is live today for all Google AI Plus, Pro, and Ultra subscribers through the Gemini app, Google Flow, and YouTube Shorts.
The model combines text, images, audio, and video as inputs to generate high-quality video output. Wired reported that the avatar feature — where you scan your own likeness and voice to insert yourself into AI-generated clips — has drawn comparisons to OpenAI's defunct Sora app.
ZDNET's David Gewirtz noted: this will help legitimate creators and increase AI-generated content on YouTube.
All Omni-generated videos carry Google's SynthID invisible watermark. OpenAI separately announced it's adopting SynthID technology for images generated through ChatGPT and its API, according to Engadget — a notable industry alignment on AI content provenance. Google is also expanding SynthID detection to Chrome and Circle to Search, allowing users to flag AI-generated images directly in the browser.
Other Changes in the Rollout
The storage reduction — Ultra users dropped from 30TB to 20TB at the new $200 price — merited little attention. You're paying $50 less but getting 10TB less storage.
Antigravity 2.0 shipped as a standalone desktop app with its own CLI tool and SDK, according to TechCrunch. Google is asking all users of the old Gemini CLI to migrate to Antigravity CLI. That's a forced transition affecting developers mid-workflow.
Google's CodeMender — its answer to Anthropic's Mythos cybersecurity model — is being opened to select external API testers, per The Verge. DeepMind CTO Kavukcuoglu confirmed Google is in talks with governments and enterprises about auditing their codebases. Pichai acknowledged Anthropic for "showing the value" of large models in security use cases.
What Comes Next
Google delivered substantial product announcements at I/O. For regular users: AI Ultra is now $100 cheaper, Gemini Spark reaches Ultra subscribers next week, and Gemini Omni video generation is available immediately.
The agent arms race moved from concept to implementation. It's in your inbox, running on Google's servers, whether your laptop is open or not.
Sources used for this briefing
This briefing was written by UBH's AI agent — these are the reporting inputs it draws on, linked so you can verify.