Gemini Omni in August 2026: API Private Beta, Omni Pro Teaser, and Latest Feature Drop

Three months after Google I/O

Gemini Omni launched at Google I/O on May 19, 2026. Three months on, the picture has shifted materially: the developer API is no longer a vague promise, Gemini Omni Pro has a concrete preview, and Google Flow has picked up the creative controls power users have been asking for since day one.

This post is the August 2026 status check — API access, the Pro tier teaser, Flow and Gemini app upgrades, and the updated pricing model heading into the second half of the year.

Developer API status: private beta is live

The biggest engineering news of the summer: Google opened a private beta for the Gemini Omni Flash API through Google AI Studio and Vertex AI. This is not a public GA rollout yet — access is limited to enterprise partners and approved developers who join the waitlist.

What beta partners get

• Endpoints: generateVideo, editVideo, and generateWithAvatar — mirroring what consumer users already have in the Gemini app and Flow.
• Rate limits: beta tiers start at 10 concurrent operations and 500 generations per day per project, with higher caps negotiable for production partners.
• SynthID + C2PA: every API response includes mandatory watermarking metadata, consistent with Google’s responsible-AI policy.

Pricing signals (not final)

Google has not published official API pricing, but beta partner documentation cites estimated rates of $0.12–$0.28 per second of generated video, depending on resolution and whether avatar mode is enabled. That lands Omni Flash competitively against Runway Gen-4 and below premium Veo 3.1 tiers for many workloads.

Analysts who modeled I/O launch assumptions at $0.10–$0.30/sec now have real numbers to work with — though Google warns these figures may change before GA.

Waitlist and timeline

Individual developers can apply through the Gemini API waitlist. Enterprise teams should route through their Google Cloud account manager for Vertex AI access. Google targets public GA in Fall 2026, with broader rate limits and self-serve billing.

Omni API vs Veo 3.1 today

Factor	Veo 3.1 API	Omni Flash API (beta)
Availability	Generally available	Private beta only
Multi-turn editing	Regenerate from scratch	Native conversational edit
Avatar support	Not available	Built-in
Max clip length	Up to 8 seconds	Up to 10 seconds (Flash)
Pricing	Published ($0.15–$0.40/sec)	Estimated ($0.12–$0.28/sec)

Bottom line for builders: if you need production API access today, Veo 3.1 remains the supported path. If you can wait until Fall 2026 — or qualify for the beta — start architecting behind a provider interface now. Our API developer guide covers the migration pattern in detail.

Gemini Omni Pro: first look

Google used August to tease Gemini Omni Pro — the higher-capability tier that was announced at I/O but had no timeline. Early preview access is rolling out to AI Ultra subscribers and select enterprise beta partners.

What Pro adds over Flash

Capability	Omni Flash	Omni Pro (preview)
Max clip length	10 seconds	Up to 30 seconds
Resolution	720p	1080p, 4K in testing
Audio	Synced native audio	Native audio synthesis with multi-role voice syncing
Character consistency	Good across turns	Stronger identity lock across longer sequences

The headline demo at Google’s August partner briefing showed a 30-second narrative clip with three distinct speaking roles — each with its own voice timbre, lip-synced to generated dialogue — without post-production audio mixing. That is the clearest differentiation from Flash: Pro is built for story-length content, not just short-form hooks.

Pro preview is not available via API yet. Google says API endpoints for Pro will follow the Flash GA rollout, likely in late 2026.

Google Flow and Gemini app enhancements

August brought three consumer-facing upgrades that matter for everyday creators:

Timeline controls in Google Flow

Flow now exposes a visual scene timeline — drag to reorder shots, trim individual segments, and preview the full sequence before export. Previously, scene ordering was chat-driven only; the timeline gives editors the precision of a traditional NLE without leaving the Omni workflow.

Avatar character consistency

The personal AI Avatar feature gained cross-scene identity lock: your digital likeness now maintains appearance, wardrobe and voice timbre across an entire multi-shot project, not just within a single generation. This closes the biggest gap between Omni avatars and professional deepfake pipelines.

Conversational voice tuning

Inside the Gemini app, you can now adjust voice characteristics mid-conversation — “make the narrator sound warmer” or “give the character a British accent” — and Omni applies the change to the next generation without breaking scene context.

These three features are live for AI Plus, Pro and Ultra subscribers on web and mobile.

Updated pricing and compute-based quotas

Google clarified the H2 2026 subscription structure alongside the August update:

Tier	Monthly	Omni access	Compute quota
AI Plus	$7.99	Flash in Gemini app + Flow	Entry limits, ~5-hour refresh
AI Pro	$19.99	Higher Flash limits + Flow credits	~3× Plus, ~5-hour refresh
AI Ultra	$100	Flash + Pro preview, 5× Pro limits	Highest consumer tier

The compute-based quota model — introduced at I/O — is now fully documented: your limits depend on prompt complexity, clip length, and features used (avatar, multi-turn edit, etc.), refreshing roughly every five hours up to a weekly cap. Heavy Flow users on Plus may hit limits faster than light Gemini app users on Pro, so tier selection should match your primary surface.

YouTube Shorts Remix and YouTube Create remain free for users 18+, unchanged from launch.

Summary and Fall 2026 roadmap

August 2026 marks the transition from “Omni is a consumer product” to “Omni is a platform.” The private API beta, Pro preview, and Flow timeline controls are the three pillars of that shift.

What to expect this fall:

1. Omni Flash API public GA — self-serve access via Gemini API and Vertex AI, with published pricing.
2. Omni Pro wider rollout — beyond Ultra preview to Pro subscribers, with 30-second clips and 1080p as defaults.
3. Image and audio output modalities — still committed, still undated, but Google confirmed internal testing is underway.
4. Workspace integration — staged rollout of Omni inside Google Docs, Slides and Vids for enterprise customers.

For the full launch story, see our May 2026 release notes and June status update. For API architecture, see the developer guide; for the Veo comparison, see Gemini Omni vs Veo 3.1.