AI Moments Newsletter – 2026-wk46

This week: ChatGPT Images 2.6 makes GIFs and generates images 50% faster, Gemini in Google Forms now analyses your responses and suggests next steps, and Gemini Omni 1.2 Flash makes 15-second video clips, 30% cheaper, and lands in Google Vids.

Research Period: 2 November to 9 November 2026

Please review the slide version of the newsletter here.


1. ChatGPT Images 2.6 makes GIFs and generates images 50% faster

What changed

We covered Images 2.5 in week 38. This time the headline is animation. OpenAI released Images 2.6 on 4 November, and ChatGPT can now turn a text prompt, a photo or even a rough sketch into a GIF. OpenAI says images generate 50% faster. Developers get two new API models: GPT-Image-2.6 Flare, which OpenAI says can produce an image in as little as six seconds at low quality, and Sunburst for more detailed work.

Why it matters

A short looping animation used to mean a designer or a video tool. Now it’s a chat message. OpenAI states Images 2.6 is available globally in ChatGPT and Codex across all subscription tiers, on desktop, mobile and web. Try it on your own account before planning a campaign around it.

Use cases

  • Animate your logo or mascot into a looping GIF for email signatures and social posts
  • Turn an existing product photo into a short animated loop for a newsletter header
  • Turn a rough hand-drawn sketch into an animated character for a social post
  • Create reaction GIFs for your customer community or team chat
  • Compare Flare and Sunburst in the API on a typical product image before automating anything

Sources: OpenAI | Superhuman


2. Gemini in Google Forms now analyses your responses and suggests next steps

What changed

Google announced on 5 November that Gemini can generate insights from the Responses tab in Forms. It reports trends, patterns across questions and suggested next steps. The counts and percentages come from Python code that Gemini runs on your actual response data, rather than from the model guessing.

Why it matters

Most of us run a customer or staff survey, glance at the charts, then never get round to the analysis. This does the first pass in minutes, not an afternoon in a spreadsheet. It’s on Business Standard and Plus, Enterprise Standard and Plus, Education Plus, and Google AI Pro and Ultra. Rapid Release domains started on 2 November. Scheduled Release domains start on 19 November, and it can take up to 15 days to appear. Users also need smart features and personalisation turned on.

Use cases

  • Check in the Admin console that smart features and personalisation are on for your team
  • Generate insights on your last customer satisfaction survey and compare them with your own reading
  • Spot patterns across questions, such as which customer group rates delivery lowest
  • Use the suggested next steps as a starting agenda for your next team meeting
  • Check the key counts against the linked Sheet before quoting them to anyone

Sources: Google Workspace Updates


3. Gemini Omni 1.2 Flash makes 15-second video clips, 30% cheaper, and lands in Google Vids

What changed

In week 36 we covered Omni 1.1 Flash. Google announced Omni 1.2 Flash on 5 November, with clips up to 15 seconds and edits by conversation, such as changing a background or lighting without starting again. API pricing is about $0.07 per second at 720p, 30% below Omni 1.1. Draft at 720p and upscale to 1080p or 4K later.

Why it matters

By our arithmetic, a full 15-second 720p clip costs about $1 through the API. Most members won’t need the API: it’s in Google Vids for Business Starter, Standard and Plus, Enterprise, Education Plus, and Google AI Pro and Ultra, rolling out from 5 November and visible within one to three days. In the Gemini app it’s Google AI Ultra only. One UK caveat: at the July launch, editing uploaded non-AI videos with Omni was unavailable in the UK and EEA. Generating clips was not restricted.

Use cases

  • Generate a 15-second product teaser in Vids from a short prompt and a product photo
  • Change the background or lighting of a clip by conversation instead of regenerating it
  • Draft at 720p, then upscale only the version you’ll actually publish
  • Build short explainer clips for staff training or onboarding in Vids
  • Review every clip showing your product or people before it goes public

Sources: Google Blog | Google Workspace Updates


Additional Noteworthy

Claude Personal Finance (US only): a read-only beta that links bank accounts through Plaid to answer questions about spending; US Pro, Max and Ultra subscribers only, so not for UK members yet. | TechCrunch

Cursor builds and publishes web apps: cloud agents now build a web app from a plain-language description and publish it once automated security checks pass; voice chat with agents followed a day later. | Cursor: Build Live Apps | Cursor: Voice Chat

Perplexity cuts Agent API costs and launches Teammates: Perplexity says every Agent API effort level costs less, medium by almost 90%; Teammates agents for ongoing work launch first for Max subscribers. | Perplexity: Agent API | Perplexity: Teammates


Strategic Awareness

Several threads suggest AI may stay pricier for longer. After the 4 November US midterms, Stratechery argues the main limit on AI is now political permission for data centres, not compute or power. HBR notes candidates in at least 21 races spent almost $90m on ads calling for data centre slowdowns or bans; Platformer looks at the pressure on AI labs. VentureBeat reports 56% of surveyed organisations saw an AI surcharge or an AI-justified renewal rise in the past year. Ramp data, via TechCrunch, shows token spend grew just 1.5% across August and September, then rose 8.9% from 23 September to 23 October. That makes your next renewal quote worth reading closely. Separately, HBR’s ‘The Human-in-the-Loop Fallacy’ questions putting staff on AI-checking duty; it’s paywalled and we only read the opening.

Sources: Stratechery: AI on the Ballot | HBR: AI Lost Big in the Midterms | Platformer | VentureBeat | TechCrunch | HBR: The Human-in-the-Loop Fallacy


Looking Ahead

Scheduled Release Workspace domains get Forms insights from 19 November. Thinking Machines said the weights for its Inkling 2 Small model would follow the week after its announcement, pending safety evaluations.

Sources: Google Workspace Updates | VentureBeat


This briefing focuses on AI developments relevant to SME operations, emphasising practical applications in productivity, cost efficiency, and business capability. For questions or to suggest sources reach out on the group WhatsApp.

Keep exploring

More from the Guild.

AI Moments lands weekly. The Guild’s sessions and articles sit alongside it and go deeper on the things that matter.