AI
“Google has launched Gemini Omni 1.1 Flash, an upgraded generative video model now available through the Gemini API and Google AI Studio, offering features including first/last frame video generation, faster/cheaper 360p draft previews, and upscaling to 1080p or 4K.”
Plain restatementGoogle released a video generation model named Gemini Omni 1.1 Flash, accessible via the Gemini API and Google AI Studio, with first-and-last-frame interpolation, a lower-cost 360p draft mode, and 1080p/4K upscaling.
This one holds up. Google did announce Gemini Omni 1.1 Flash on August 27, 2026, and it is available to developers through the Gemini API and Google AI Studio, as well as to paid Google AI subscribers in Flow. The specific features listed in the post are all documented on Google's own developer pages: generating video between a chosen first and last frame, a cheap 360p draft mode, upscaling to 1080p or 4K, scene extension, and using short video clips as references. Two small things are worth knowing. The 1080p and 4K outputs are produced by upscaling, not generated natively at those resolutions, and the post's headline blurs that even though its caption gets it right. Also, the "up to 60% faster and one third the cost" figures come from Google itself, not from any independent test, though Google's published billing table does confirm the one-third price ratio. The headline also garbles the name as "Gemini 1.1 Omni" rather than Gemini Omni 1.1 Flash.
Google [drifted from the evidence:] has launched Gemini Omni 1.1 Flash, [drifted from the evidence:] an upgraded generative video model now available through the Gemini API and Google AI Studio, [drifted from the evidence:] offering features including first/last frame video generation, faster/cheaper 360p draft [drifted from the evidence:] previews, and upscaling [drifted from the evidence:] to 1080p or 4K.
Google [added by the neutral restatement:] released a video generation model named Gemini Omni 1.1 Flash, [added by the neutral restatement:] accessible via the Gemini API and Google AI Studio, [added by the neutral restatement:] with first-and-last-frame interpolation, a lower-cost 360p draft [added by the neutral restatement:] mode, and [added by the neutral restatement:] 1080p/4K upscaling.
Red-tinted words in the claim drifted from the evidence. Green-tinted words are what a neutral restatement needs.
The trace / claim to source
- The model exists, is named Gemini Omni 1.1 Flash, and was announced by Google on 2026-08-27. This is confirmed on Google's blog, developer docs, DeepMind model card, and official X account.
- It is available through the Gemini API and Google AI Studio, exactly as the caption states.
- First-and-last-frame generation is real and is documented in the API changelog as interpolation via `image_to_video` with up to two images.
- The 360p draft mode is real, and Google's own claim is "up to 60% faster" at "a third of the cost" versus 720p. The one-third cost ratio is corroborated by Google's published token-per-second billing table.
- 1080p and 4K upscaling of finished video is real and documented.
- The caption's other listed features check out: scene extension for longer videos, and short video clips as references. Google's X post specifies up to three seconds of reference video.
- Omitted qualifier: the post's headline slide reads "up to 4k video generations," but Google's API documentation labels 1080p and 4K as upscaled output rather than native generation. The post's own later slide corrects this by saying it "can turn finished videos into" 1080p or 4K versions, and the caption says "Upscale," so the imprecision is confined to the headline and is undone within the same post.
- Minor naming error: the headline slide calls the model "Gemini 1.1 Omni," dropping "Flash" and reversing the word order. The correct name, used in the post's own caption, is Gemini Omni 1.1 Flash. This does not create version ambiguity, since the API model ID `gemini-omni-1.1-flash` matches unambiguously.
- Marketing as evidence: the "60% faster" and "one-third the cost" figures are Google's own, not independently measured. The cost ratio is verifiable from Google's billing table, but the speed figure carries an asterisk on Google's own page and has no independent benchmark behind it.
- The precise lifecycle designation on the Gemini API surface. Google calls it "production-ready" and the model ID lacks a preview suffix, but the Enterprise Agent Platform page labels the same model Preview under Pre-GA terms. The post says "launched" and "available," which is supported either way, so this does not change the verdict.
- The content of the footnote attached to Google's "up to 60% faster*" claim. I read the blog text through search-result extracts rather than a full page fetch, and the footnote body was not in the retrieved material.
- Whether the 60% speed improvement holds in ordinary use. No independent measurement was found, and the claim is one day old.
- I read all official pages via search-result extracts of those pages rather than direct fetches. The extracts were substantive and mutually consistent across four separate Google properties, so I regard the release facts as established, but I did not read any page in full.
Every element of the claim appears on Google's own channels. Google's blog post introduces Gemini Omni 1.1 Flash as a production-ready update for developers offering improved control over generative video, with scene extension, start and end frame specification, and 4K output, accessible through Google AI Studio or the Gemini Enterprise Agent Platform. On the draft mode, Google's own wording is: "Generate lightweight previews in 360p resolution up to 60% faster* and at a third of the cost compared to Omni 1.1's standard 720p resolution." Note the asterisk, which points to a footnote qualification on Google's page. The developer documentation confirms the model identifier and description: Gemini Omni Flash (gemini-omni-1.1-flash) is described as a high-performance multimodal model designed for high-speed video generation, editing, and cinematic control. The changelog records the specific features: interpolation using first and last frames via the image_to_video task with up to two images, and a new resolution parameter in video_config supporting 360p, 720p (default), 1080p, and 4k outputs, with 1080p and 4K outputs generated using upscaling; the existing gemini-omni-flash-preview endpoint is to be deprecated on September 30, 2026. Google's official X account lists the same feature set and states where it is rolling out. Google says the model is rolling out in Google AI Studio, Flow, and the Gemini Enterprise Agent Platform. Consumer-surface availability is broader than the post mentions: Omni 1.1 is available to all Google AI Plus, Pro and Ultra subscribers globally in Google Flow, and scene extension is available to those subscribers in the Gemini app. Third-party press adds one qualification the post's headline does not carry. Neowin notes that per Google's Gemini API documentation, both 1080p and 4K are upscaled rather than generated natively, with 720p remaining the default and 360p also available. Another outlet makes the same point directly, noting that 4K is not described as native generation and that the API documentation explicitly labels it as upscaled output. Pricing, from Google's own pricing page: Gemini Omni Flash charges 1931 tokens per second of 360p video, 5792 tokens per second of 720p, 8688 for 1080p, and 17376 for 4k video outputs with audio. That 1931 versus 5792 ratio is almost exactly one third, which independently corroborates the "one-third the cost" claim from the vendor's own billing table rather than only from its marketing copy.
Complete reasoning
The reply is formatted for pasting into the thread where the claim is circulating.
Compact share page: ai.trueseeker.com/s/6200b662ac37/-kW9hCLpFl59d1GtZZINNUZ6CFD
Ask this case
Answers come only from the case file above; nothing is added.
Is Gemini Omni 1.1 Flash a real product?
Yes. Google announced it on August 27, 2026, and it is confirmed on Google's blog, developer documentation, DeepMind model card, and official X account.
Where can people actually use it?
It is available through the Gemini API and Google AI Studio, as stated in the claim, and it is also rolling out in Flow and the Gemini Enterprise Agent Platform.
Does it really generate 4K video?
Not natively. Google's own API documentation says 1080p and 4K outputs are produced by upscaling finished video, not generated directly at those resolutions. The post's headline slide blurs this, but a later slide and the caption describe it correctly as upscaling.
Are the speed and cost claims verified by anyone outside Google?
The one-third cost figure is backed up by Google's own published billing table, which shows a near-exact one-third ratio between 360p and 720p pricing. The 60% faster claim comes only from Google's marketing copy and carries an unexplained footnote, with no independent test found.
Is the model name in the post correct?
The caption uses the correct name, Gemini Omni 1.1 Flash, but the headline slide calls it 'Gemini 1.1 Omni,' which drops a word and reverses the order. This is a minor error and does not cause confusion since the API model ID matches the correct name.