Gemini Omni Release Date

Gemini Omni is one of the two AI launches everyone is watching for at Google I/O 2026. As of May 18, 2026, Google has not officially confirmed the model or a release date, but the leak coverage and the I/O schedule point to the same answer: May 19, 2026.

What is officially confirmed

Two solid facts:

1. Google I/O 2026 runs May 19-20, 2026. The main keynote starts at 1 PM ET on May 19. Google has confirmed the dates and the keynote slot.
2. The name "Omni" appears inside the Gemini app. UI strings referencing an Omni video model surfaced in late April and early May 2026 and have been reproduced by multiple analysts. This is the primary evidence that Omni is a real internal Google project.

Beyond those two points, everything else is leak-driven and should be treated as expectation rather than confirmed.

Leak timeline

The Omni leaks have arrived in waves:

• Mid-April 2026. First UI string spotted inside the Gemini app referencing an Omni model selection.
• Late April 2026. Reports describing Omni as a unified multimodal pipeline that handles text, image, video, and audio in one model.
• Early May 2026. Reporting that Omni may replace or supplement the Veo 3.1 pipeline inside the Gemini app, with one source citing a 2-hour 1080p clip capability.
• May 12, 2026. Google's pre-I/O Android Show announced Gemini Intelligence and Googlebooks, but did not mention Omni — keeping the name held back for the main I/O keynote.
• May 18, 2026. Day before I/O. No official acknowledgment of Omni from Google yet.

Each wave has been independently sourced, and the description has stayed consistent across coverage. That consistency is the main reason Omni is treated as a real product rather than a single rumor.

Realistic launch scenarios

Three plausible scenarios for what happens on May 19:

| Scenario | Probability | What it means | |---|---|---| | Omni announced and rolled out inside the Gemini app on May 19 | High | Day-one availability for personal Google accounts | | Omni announced as a preview, broader rollout over June | Medium | Capacity-constrained launch; matches some hardware-heavy launches | | Omni folded into Veo 4 or a new Gemini model number, name dropped | Low | Possible if Google reorganizes the brand at I/O |

The first scenario fits Google's recent pattern with consumer AI features in the Gemini app: announce at I/O, ship inside the app the same week. Pre-I/O Android Show launches like Gemini Intelligence have already shipped or are rolling out, suggesting Google has the capacity to launch big at this I/O.

Day-one availability expectations

Based on how Google has launched recent Gemini features:

• Inside the Gemini app on web and mobile for personal Google accounts, possibly gated by region
• Quota-limited free generations with paid Google AI Pro ($7.99/month) and AI Ultra unlocking more
• No immediate Vertex AI exposure, since the leaks consistently place Omni inside the Gemini app rather than as an enterprise product
• Gemini API access likely later in 2026 once the Gemini app surface stabilizes

If the unified multimodal architecture holds, Omni's launch will mark the first widely deployed model from a major lab that natively outputs across text, image, video, and audio. That is a significant moment regardless of how the day-one experience compares to specialized models.

How to track the launch

Reliable real-time sources on May 19:

• Google I/O 2026 livestream for the keynote announcement
• Google blog and DeepMind blog for technical details
• The Gemini app itself — the new model may simply appear in the model selector after the keynote
• Major tech outlet liveblogs, including Android Central's I/O 2026 live blog

For deeper context on what Omni is expected to do, see What Is Gemini Omni?. For the parallel Veo 4 launch track, see Veo 4 release date and the Veo 4 vs Gemini Omni comparison.

What to use today while you wait

If you need cross-modal AI generation before Omni lands:

• Gemini 3.1 Pro for text and reasoning, with image generation through Nano Banana Pro
• Veo 3.1 for video generation, free for personal Google accounts at 10 generations per month
• Cross-platform creative workflows like Elser.ai for image-to-video and animation that work across providers

For interactive 3D content, which Omni does not produce, see What Is Happy Oyster? and Happy Oyster vs Gemini Omni.