Google is bringing Gemini Omni editing and a reusable face-and-voice avatar to Vids, but the sharper distinction is account-level identity across Vids and Gemini rather than a new category of AI video; Vids already offered Veo generation and customizable avatars, while specialist rivals already sell digital presenters.
Google's most consequential Vids update is not another synthetic presenter. It is a reusable version of the account holder's face and voice that can move between Vids and Gemini, coupled with a video model that accepts revisions through conversation.
That distinction matters because Google is entering an established field. Vids already generated clips and offered customizable AI avatars, while specialist services already sell avatar video. The July release instead combines personal likeness, account distribution and Omni editing inside products Google already supplies to consumers and organizations.
Google announced two additions on July 16: Gemini Omni in Vids, and a personal avatar made from the user's face and voice. Omni can generate a clip from text and reference images, then take instructions to change a background, correct lighting or add effects. Google says the edits can be made step by step rather than by rebuilding the clip from the start.
Neither AI video nor avatars are new to Vids. Google introduced the service in 2024 as a Workspace application alongside Docs, Sheets and Slides, with generated storyboards, suggested stock media and preset voiceovers in the original product description. In August 2025, it announced a no-cost consumer editor without AI capabilities while reserving new generative features, including AI avatars, for paying customers.
By April 2026, Google said in another update that every personal Google Account received 10 free Veo 3.1 generations a month. That release also described paid avatars that could be placed in scenes, directed to interact with uploaded objects, and modified with new clothes or backgrounds while retaining a consistent voice and identity.
The comparison base is therefore narrower than Google's launch language suggests. July adds the account holder as an avatar and brings Omni's conversational workflow into Vids. It does not introduce AI-generated video, free generation or directable avatars to the product.
Omni is also older than the Vids integration. Google introduced the model in May with support for text, image, video and audio inputs and conversational editing. At that point, Omni Flash was launching in the Gemini app and Google Flow for subscribers, with no-cost access planned through YouTube Shorts and YouTube Create, according to contemporaneous coverage.
Google says the personal avatar is linked to the user's Google Account and restricted to the account holder's likeness. Its Vids setup guide shows that creation is more involved than attaching a selfie to a prompt: an adult user scans a QR code with a phone or tablet, then follows prompts to record a face and voice. The instructions say the account owner must complete the process, minors must not record themselves, and no other faces or voices should appear in the background.
Once created, the avatar can be selected as an Omni ingredient in Vids, given a script and prompt, and used in subsequent edit, extend or recreate actions. In Gemini, the same avatar appears as an @ mention of the Google username and can be used in Omni videos or Nano Banana images, according to the account instructions.
Retaking or deleting the recording changes the avatar data available to services connected to the account. Google says deletion removes the recorded selfie and voice from its systems and prevents new content from being made with that avatar. It does not delete videos or other material already created and published.
That cross-product reuse is Google's clearest distribution advantage, not a technical monopoly on digital presenters. A comparison of the release identifies HeyGen, Synthesia, Captions and D-ID as existing avatar-video alternatives. OpenAI's Sora had also allowed people to place themselves in generated video before the app shut down.
The archived sources do not provide like-for-like prices, generation costs or output-quality tests across those services. They support a distribution argument—Vids sits in Workspace and the avatar also works in Gemini—not a conclusion that Google's output is cheaper or better.
The conversational interface does not eliminate the product's technical boundaries. Google's current Vids documentation specifies 720p output at 24 frames per second, in either 16:9 landscape or 9:16 portrait. New clips can use up to seven reference images. An edit can use up to three images or an avatar with an uploaded clip, but when the upload exceeds 10 seconds, Vids edits only its final 10 seconds.
Most Vids users can generate up to 50 videos each month, with the allowance resetting at midnight Pacific time on the first day of the month. For Pro and Ultra accounts using family sharing, that monthly capacity is shared. Session history also disappears when the Vids tab closes unless a generated clip has been inserted into the project.
The Gemini app offers a related but different workflow. It accepts one video and up to five images in a prompt, supports repeated editing turns, and can be asked to generate audio with video. The finished result can be downloaded or sent to YouTube, according to Google's Gemini instructions.
These are useful creation tools, but the documented limits make Omni a short-form generation and revision system rather than an unrestricted replacement for a conventional editing stack. The sources contain demonstrations and company descriptions, not independent evidence that repeated edits reliably preserve a face, voice, object or scene.
Google says Gemini Omni and personal avatars in Vids are available to Google AI Pro and Ultra subscribers and eligible Workspace customers. The help pages narrow that further: avatar ingredients are English-only, personal AI Pro and Ultra access in Vids is listed as U.S.-only, and supported Workspace editions span specified business, enterprise, education, nonprofit and Essentials accounts.
Personal avatars require users to be at least 18 and are unavailable in the European Economic Area, Switzerland and the United Kingdom. Uploading a video for Omni editing is also unavailable in those regions and in Illinois and Texas; affected Vids users may edit only AI-generated clips retained in their generation history.
The eligibility record is not fully consistent. One account of the launch says an avatar feature had previously become available to all U.S. Google Account holders. Google's current help pages place personal-avatar creation behind eligible paid AI or Workspace plans. The archived material does not establish whether the earlier feature was the same account-bound personal avatar or one of Vids' pre-existing avatar tools.
Every generated clip receives an invisible SynthID watermark. Google says this permits verification that AI created the video, but an imperceptible marker is not an on-screen disclosure to viewers. The account restriction likewise governs creation within Google's workflow; it cannot retract avatar videos that have already been published.
The unresolved question is whether account integration is enough to make Google's personal avatar more useful than established alternatives. The sources do not contain comparative prices, rendering times, independent quality tests or evidence about identity consistency over repeated Omni edits. Those measurements—not the existence of another avatar generator—would show whether Vids changes production economics for training, product demonstrations and internal communications.
Organizations considering the tool still face a policy decision: who may create synthetic versions of employees, for which uses, and under what retention rules. The product documentation establishes individual account and deletion controls, but it does not answer those workplace-governance questions.
For now, Google has connected three assets it already controlled: Workspace distribution, the Gemini creation interface and the user's account. The release becomes strategically significant if that combination produces repeatable, economical video work within the documented quotas and regional rules. The available evidence establishes the integration; it does not yet establish the result.
OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
RoboTTT extends a robot policy’s visuomotor context to 8,000 timesteps and raises its average rubric score from 42% to 79%, but the evidence comes from three author-run assembly tasks and the longest task was completed in only two of ten trials.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Nebius has arranged its first senior secured facility against an operating GPU deployment and one customer’s cash flows. The deal adds project-level debt to its expansion toolkit, but an unnamed customer and a larger rival financing limit what it proves about the rest of Nebius’s backlog.
Cursor says automatically running a repository-root git.exe on Windows does not meet its criteria for patching, while the researcher calls it an untrusted-search-path defect and separate research shows the same weakness across several competing AI coding tools.
Huawei publicly displayed a 16-cabinet Atlas 950 configuration rated at 1 EFLOPS in FP8, providing tangible evidence of its system-scale AI strategy while leaving price, power use and sustained workload performance undisclosed ahead of the full system's planned fourth-quarter release.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
UK testing places leading open-weight models four to seven months behind selected closed-model cyber results, yet longer attack chains, U.S. benchmarks and mixed cost comparisons show why that interval is a warning signal rather than a universal capability clock.
TSMC reached the top of its second-quarter revenue guidance and raised its 2026 outlook, but a one-off investment gain boosted profit growth while the company committed more capital to 2-nanometer production, advanced packaging and an undated Arizona expansion.
Moonshot AI's Kimi K3 added to a global technology selloff with near-frontier performance, but unreleased weights, mixed cost comparisons and a recommended 64-accelerator deployment leave its effect on chip demand unresolved.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
Two binding EU decisions require Google to give rival AI services comparable access to 11 Android features and offer eligible search competitors a restricted, anonymized dataset, but phased deadlines, certification, pricing and privacy safeguards leave the competitive effect unproven.
Databricks has signed a term sheet for a Coatue-led financing at a $188 billion valuation, while unidentified sources put the round at $3 billion. The proposed capital would deepen its push into AI governance, data agents and operational databases, but the transaction remains open and the company supplied no new operating figures.
Netflix says generative AI workflows were used on roughly 300 titles in 2026, but its only quantified example tied to that disclosure covers 17 minutes and does not establish how much finished material, spending or labor changed across the slate.
Gemini 3.5 Pro missed its expected June rollout. An anonymous-source account says a late-June data change fell short of Google's coding goals, but Google has confirmed only partner testing—not the reported cause, a new date, or public results and pricing.
Intel plans to adopt Gemini Enterprise across engineering, supply-chain and corporate work while extending chip-simulation capacity into Google Cloud, but the companies supplied no rollout schedule, contract value or performance baseline.
The Navy has approved an immediate department-wide framework for turning data into operational effects, but the public rollout leaves budgets, owners, deadlines, performance measures and assurance rules to a promised implementation roadmap.
Runta has raised a $20 million seed round led by Andreessen Horowitz for an execution layer that governs agents at the operating-system and network levels, entering a market where cloud runtimes and sandbox providers already offer overlapping controls.