Gemini Omni 1.2 Flash generates longer AI videos at a lower per-second price, while OpenAI is adding animated GIFs and scalable graphics export to its image tools. The updates affect anyone creating short product clips, social media posts, presentation graphics, or simple animations. They are particularly relevant because revisions can take less time and not every draft needs to be rendered at the highest quality.
What is new in image and video generation?
Generative Artificial Intelligence (AI), meaning software that creates new content from text or reference material, is becoming faster and more flexible in both product families. Google is introducing Gemini Omni 1.2 Flash with videos up to 15 seconds long. Its Omni 1.1 predecessor was limited to ten seconds.
According to Google, you can set the video length in one-second increments from three to 15 seconds. The model also supports conversational editing: You can ask it to change the background, lighting, or aspect ratio of an existing clip instead of generating the entire video again. Google uses the term “Conversational Editing” for these step-by-step revisions through follow-up instructions.
Google also claims stronger adherence to reference material. In practice, this means people, clothing, product packaging, or brand colors should remain more similar across scenes and camera angles. One example provided by the company is a virtual try-on clip that combines a model photo and images of three outfits into a 15-second sequence with 360-degree turns.
OpenAI, meanwhile, is expanding Images 2.6 across ChatGPT, Codex, and its application programming interface (API). An API is a technical connection that lets other applications use a service. According to a summary of the Images 2.6 changes, image generation is supposed to be 50 percent faster than before, although no independent benchmark results were published.
In ChatGPT, you can turn a text prompt or an uploaded photo into an animated GIF. GIF stands for Graphics Interchange Format, a common file format for short image animations. You can adjust the animation speed and edit individual frames; a dancing cat was the everyday example shown with the feature.
OpenAI also lists Scalable Vector Graphics (SVG) export, a format that can be resized without visible blurring and is commonly used for icons, diagrams, and interface elements. However, the available sources associate SVG export with the Image API. They do not establish that you can already produce SVG files directly in the regular ChatGPT interface as easily as GIFs.
How can you reduce time and cost?
According to Google, Gemini Omni 1.2 Flash costs $0.07 per second for video at 720p resolution, 30 percent less than Omni 1.1. At that listed rate, a full 15-second generation works out to $1.05. MarkTechPost reports the same per-second price and reduction, but the available short summary does not provide independent verification.
The lower rate is only one part of the intended workflow. Google suggests creating early drafts at 360p or 720p, where “p” refers to the vertical image resolution. You can then upscale only the selected final version to 1080p or 4K, a substantially higher resolution.
Conversational editing may reduce unnecessary generation as well. If only the background or lighting is wrong, you should be able to revise that detail rather than pay to start with a completely new version. Whether these changes work consistently in every case has not been independently verified; claims about fewer visual jumps, better lip synchronization, and more stable characters come from Google.
For OpenAI, the main time saving comes from faster drafts. GPT-Image-2.6 Flare is said to return an image in as little as six seconds at low quality, while Sunburst is aimed at more detailed compositions and greater consistency during editing. However, the supplied sources contain no full per-image price table for Images 2.6, preventing a direct cost comparison with earlier versions.
A new complexity control determines how much planning the model performs before rendering, meaning before it produces the final image. OpenAI says a low setting suits focused tasks such as replacing a background or changing a sign. A high setting is intended for dense text and complex layouts, but it may take longer and introduce unwanted changes.
How can you try the features in practice?
Step 1: Choose the right format
First decide whether you need a single image, a short GIF animation, or a video clip. A moving reaction image or simple product animation may only require a GIF in ChatGPT. For a continuous scene involving camera movement, dialog, or several outfits, Gemini Omni 1.2 Flash is the more suitable of the tools described here.
Step 2: Start with a small draft
With Gemini, specify only the duration you actually need between three and 15 seconds, then begin at 360p or 720p. With OpenAI, you can use low complexity for a quick and narrowly defined image draft. Dense labels or complicated relationships among several objects are more appropriate for a higher setting.
Step 3: Revise specific details
Avoid describing the entire result again when only one detail is wrong. In Gemini, for example, you can request a different background, revised lighting, or another aspect ratio. For a GIF, you can adjust the animation speed and edit individual frames, the separate images that make up the animation.
- Describe the subject, intended motion, duration, and output format as specifically as possible.
- Check a lower-cost video draft for visual jumps, character consistency, and text before upscaling it.
- Use follow-up instructions to change individual details rather than immediately starting a new draft.
- Increase resolution or complexity only for the selected final version.
One practical workplace example is Google’s virtual try-on clip built from a model photo and three product images. A second comes from OpenAI’s features: A social media team can turn a photo into a GIF, adjust its speed, and correct individual frames without assembling the animation in a separate tool.
Pros and Cons of the new tools
Pros:
- Longer clips – Gemini raises native video length from ten seconds to as much as 15 seconds.
- Lower listed price – The stated price for 720p video falls by 30 percent to $0.07 per second.
- Targeted revisions – You can edit backgrounds, lighting, aspect ratios, and individual GIF frames after generation.
- More output formats – Animated GIFs and, through the OpenAI API, SVG files can remove extra production steps.
Cons:
- Provider claims – Statements about speed, reference adherence, and consistency have not been independently confirmed.
- Incomplete pricing – The sources do not provide a complete per-image cost overview for OpenAI Images 2.6.
- Limited duration – Fifteen seconds is useful for short scenes but not for longer explainers or training videos.
- Unclear interface support – SVG export is listed for the API but not clearly confirmed for regular ChatGPT use.
Rapid usage growth helps explain why providers are accelerating their image tools. OpenAI says people now generate more than four billion images per week across ChatGPT and the API, compared with more than three billion eight weeks earlier. The report on the GIF rollout describes it as an ongoing release for ChatGPT and Codex users on the web, desktop, and mobile apps. These usage figures also come from OpenAI and are not presented as an independent measurement.
What does this mean for your work and for Switzerland?
If you are just getting started, a short and clearly limited test is the most sensible first step. You could create a GIF from an existing photo or a three-to-five-second video draft at low resolution. This lets you assess motion, text, and character consistency without committing to a larger production.
If you already have experience, you can get more from the tools by using reference images and follow-up instructions systematically. With Gemini, you can separate drafting, detail correction, and upscaling; with OpenAI, you can choose complexity and quality independently. Saved reference images of people, pets, or characters are also meant to be reusable later without another upload.
According to the report on availability and editing, Google is rolling out Gemini Omni 1.2 Flash through services including Google AI Studio, Google Cloud, Flow, Google Vids, the Gemini app for AI Ultra subscribers, and Google Ads. Google also names the Gemini API and Gemini for business. All generated videos are supposed to carry an invisible SynthID watermark that identifies their origin.
The sources provide no Switzerland-specific prices, availability details, supported-language list, or data protection terms. Whether a feature has already appeared in your Swiss account therefore depends on the rollout and the service you use. For workplace content, you also need to establish whether uploaded images of people or products can be processed under your organization’s policies; the sources provide no details on this point.
These updates make short visual content less expensive, easier to revise, and simpler to animate inside ChatGPT. Gemini offers the more clearly quantified price advantage and longer clips, while OpenAI focuses on rapid image drafts, GIFs, and additional output formats. The unresolved issues are how reliably the promised consistency holds up beyond provider examples and which features are actually available across all accounts and interfaces.
Sources
- Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips – Google, 2026-11-05
- Google Releases Gemini Omni 1.2 Flash: 15-Second AI Video at $0.07 per Second – MarkTechPost, 2026-11-05
- Google veröffentlicht Gemini Omni 1.2 Flash mit längeren Clips und günstigeren Preisen – THE DECODER, 2026-11-05
- OpenAI Launches GPT-Image-2.6 Flare and Sunburst: Adds GIFs, SVG Export and 6-Second API Generation – MarkTechPost, 2026-11-05
- ChatGPT kann jetzt animierte GIFs erstellen – OpenAI rollt neue Bildfunktion aus – THE DECODER, 2026-11-04
Image: Amar Preciado via Pexels