Gemini Omni 1.5: Release Status, 15-Second Video Rumor and API Pricing

Gemini Omni 1.5 release date: what is actually confirmed?

There are three different things behind this search: the released Gemini Omni family, the stable Omni 1.1 Flash API, and the unannounced Omni 1.5 label.

Google introduced the first Omni Flash model at I/O 2026 with the positioning “create anything from any input — starting with video.” Its announcement describes combining text, images, audio and video references with Gemini's reasoning and world knowledge. It also explicitly describes image and audio output as future expansion. “Any-to-any” therefore describes the broader direction; it does not mean every output modality is publicly available today.

The Gemini API changelog confirms August 27, 2026 as the general-availability release of gemini-omni-1.1-flash. It adds video extension, first/last-frame interpolation and resolution control. That entry also marks the older gemini-omni-flash-preview endpoint for deprecation on September 30. Its continued appearance on a pricing page should not be read as a recommendation for a new integration: start with the stable ID.

The official model catalog, Omni model page, developer guide and changelog checked for this article do not establish an Omni 1.5 launch. This does not prove Google has no internal project with that name. It means there is currently no verified public release contract to build against.

Claim	Evidence available on October 19, 2026	Status
Gemini Omni is a Google video generation and editing family	Google launch announcement and DeepMind product page	Confirmed
Omni 1.1 Flash has a stable developer API	Google model page and August 27 changelog	Confirmed
Current Omni output is 3–10 seconds	Official 1.1 Flash model specification	Confirmed
Omni 1.5 is in development and generates 15 seconds	Circulating screenshot attributed to @lyraxana	Unconfirmed
Omni 1.5 has a launch date, public API ID or price	None found in the official sources checked	Not announced

Does Gemini Omni 1.5 support 15-second videos?

The 15-second claim is not an official specification. The supplied screenshot says Google is developing Omni 1.5 with 15-second output rather than the standard 10 seconds. We could verify the current 10-second maximum against Google's model page, but could not independently verify the original post, its publication date or a Google source supporting the proposed upgrade. A relative “14 hours” timestamp in a screenshot does not establish a calendar date.

If a future release raises a single generation from 10 to 15 seconds, that would be 50% more maximum duration. It could give an advertisement, short narrative or camera move more time to develop. It would not by itself demonstrate better subject consistency, physics, sound synchronization or cost efficiency.

Also distinguish single-generation duration from video extension. Omni 1.1 Flash already documents generating a 3–10-second continuation at the end of a clip. A longer edited or extended sequence is not proof that an unannounced model supports a single 15-second generation. The guide separately limits uploaded extension inputs to 10 seconds, with different handling for model-generated multi-turn history.

There is no confirmed 15-second Omni 1.5 price. Multiplying the current model's approximate per-second rate by 15 would be a hypothetical budget calculation, not a future model quotation.

What can Gemini Omni 1.1 Flash do today?

The released model is already more than a text-to-video generator. Its main practical distinction is a workflow that combines generation with conversational changes.

• Text and reference inputs: generate from a prompt, animate images or use reference material to steer a scene. The developer guide describes native multimodality across text, images, video and audio; feature availability and audio-reference restrictions depend on the documented workflow. The launch announcement initially limited audio references to voices.
• Conversational editing: change an object, environment, action or camera angle over multiple turns while asking the model to preserve other parts of the scene. Continuity is a capability to evaluate, not a guarantee that every edit preserves identity perfectly.
• First and last frames: use up to two images in the image_to_video task to guide the transition.
• Extension: append a continuation to an existing clip. This is an end-of-clip operation, not insertion into the middle of a timeline.
• Resolution: 360p drafts, 720p default, and 1080p or 4K outputs using upscaling. A 4K deliverable does not establish native 4K generation.
• Native audio and provenance: generated video includes sound and Google's invisible SynthID watermark.

The model page lists a 1,048,576-token context window, 3–10-second output and 24 FPS. Context length describes input capacity; it does not turn the model into an unlimited-duration video generator.

Uploaded-video editing and extension have regional restrictions in the EEA, Switzerland and the UK, while editing or extending the model's own generated videos is supported in available regions. The guide also has restrictions around images of minors and recognizable people. Check these conditions for your actual workflow before adopting it.

Gemini Omni API pricing: why “$0.10 per second” is an approximation

Google's published paid-tier prices for gemini-omni-1.1-flash are:

Billable item	Published price
Text, image, video and audio input	$1.50 per 1 million tokens
Text output, including thinking tokens	$9.00 per 1 million tokens
Video output	$17.50 per 1 million tokens

Google states that 720p video consumes 5,792 output tokens per second. The conversion is:

5,792 × $17.50 / 1,000,000 = $0.10136 per second of 720p video output

That makes the video-output component approximately $0.30408 for 3 seconds, $0.5068 for 5 seconds, or $1.0136 for 10 seconds. Input, text/thinking output and any other applicable charges are additional. These figures describe Google's current 1.1 Flash rate, not APIMaster pricing and not Omni 1.5 pricing. Do not reuse the 720p token conversion for every resolution without checking the current billing documentation.

The pricing page lists no free API tier for Omni. Subscription access in the Gemini app or Flow, and the rollout on YouTube surfaces, are separate products with their own eligibility and usage limits; they do not grant unlimited developer API usage.

How to use the confirmed Gemini Omni API

For the current official API, open Google AI Studio, configure a paid Gemini API project and follow the Omni developer guide. The documented integration uses Google's Interactions API with the stable model ID. It is not an OpenAI Chat Completions request.

This minimal Python example follows Google's text-to-video example. Install a current google-genai SDK and configure GEMINI_API_KEY in your environment:

import base64
from google import genai

client = genai.Client()
interaction = client.interactions.create(
    model="gemini-omni-1.1-flash",
    input=(
        "A paper boat drifting across a quiet lake at sunrise. "
        "A single continuous shot, gentle ripples, natural ambient sound."
    ),
)

with open("omni-video.mp4", "wb") as output:
    output.write(base64.b64decode(interaction.output_video.data))

The example illustrates the documented base64-delivery flow and was checked against the guide; we did not run a paid video generation for this article. Larger outputs can use URI delivery. For subsequent editing, follow previous_interaction_id and storage requirements in the guide: setting store=false prevents later edits through that stored interaction.

Do not substitute gemini-omni-1.5 into this example. No official source checked documents that ID. An APIMaster Omni availability search also does not create access when there are no public routes.

Gemini Omni vs Veo, Gemini 1.5 and available alternatives

Gemini Omni and Veo: Google presents Omni as bringing Gemini reasoning, references and conversational editing into video creation. Veo remains a separately named video family. The Omni announcement does not establish that every Veo endpoint is retired or that a Veo request can be relabeled as Omni.

Gemini Omni 1.5 and Gemini 1.5: Gemini 1.5 Pro and Flash are the earlier long-context model family associated with multimodal understanding. The shared “1.5” number does not make them the rumored Omni video model. Similarly, Kling's use of “Omni” does not establish a relationship with Google's family.

Available APIMaster alternatives: if the immediate requirement is a video API through APIMaster, compare Seedance 2.5 and Seedance 2.0 for your quality-sensitive shots, then test Fast and Mini when reducing experimentation cost matters. The checked 720p starting rates are $0.219541, $0.143625, $0.072576 and $0.0378 per second respectively. These are independent models with different behavior and request fields.

For a first alternative-model test, register, create an API key, and use https://apimaster.ai/v1 with the Seedance 2.5 API documentation. Submit a short task, poll its returned task ID and inspect the completed video before sending a batch. A task ID confirms submission, not a successful generation. Compare cost per accepted clip alongside identity consistency, motion, audio and latency.

FAQ

Has Gemini Omni 1.5 been released?

No public release was found in the Google sources checked on October 19, 2026. The documented stable Omni API is gemini-omni-1.1-flash. An unconfirmed development claim does not establish public availability.

Does Gemini Omni 1.5 generate 15-second videos?

A circulating screenshot makes that claim, but it is not a verified Google specification. The current official 1.1 Flash model page lists 3–10-second output. Video extension is a separate capability.

What is the Gemini Omni 1.5 release date and model ID?

Neither is confirmed in the official sources checked. gemini-omni-1.5 in a search or third-party listing should not be treated as an official API identifier.

What does the Gemini Omni API cost?

For the released 1.1 Flash model, Google lists $1.50 per million input tokens, $9.00 per million text-output tokens and $17.50 per million video-output tokens. At 5,792 video tokens per 720p second, video output alone is approximately $0.10136/s. Omni 1.5 has no confirmed price.

Can I buy Gemini Omni through APIMaster?

No public active route was found for Omni 1.1 Flash or the Omni 1.5 search label in the October 19 marketplace check. Check the model card for changes. Seedance is available as a separate alternative, not as an Omni route.

Can Gemini Omni output native 4K video?

The current model supports 4K output through upscaling. Google's documentation distinguishes 360p and 720p generation from 1080p and 4K upscaled deliverables.

Is Gemini Omni 1.5 the same as Gemini 1.5 Pro or Flash?

No. Gemini 1.5 Pro and Flash belong to an earlier long-context multimodal model family. The unannounced Omni 1.5 label refers to a claimed future video-generation model.

Sources and next steps

Checked October 19, 2026:

• Google: Introducing Gemini Omni — family positioning, conversational editing, rollout and SynthID.
• Google DeepMind: Gemini Omni — product capabilities and generation versus upscaling.
• Google Gemini API model catalog and Omni Flash model specification — stable ID, duration, frame rate and context limit.
• Google Gemini API changelog — August 27 GA release and preview deprecation notice.
• Google Omni developer guide — Interactions API, extension, editing and regional restrictions.
• Google Gemini API pricing — token prices and 720p conversion.
• APIMaster public marketplace endpoint /api/marketplace-data?model=<id> — active routes and observed prices. The 15-second claim comes from a supplied screenshot attributed to @lyraxana; its original post and date were not independently verified.

Need a video API today? Create an APIMaster account, create a key, and use https://apimaster.ai/v1 with seedance-2.5 or another model whose card lists active routes. The checked Seedance Mini rate is up to ~50% below its marketplace reference price; compare the current cards before choosing. For Google Omni itself, use the confirmed gemini-omni-1.1-flash Google API while waiting for any official 1.5 announcement.