HAI Daily

Gemini Omni 1.2 Flash Video, OpenAI's $11 Billion Close and China's New Lithography Tools

Google's Gemini Omni 1.2 Flash makes 15-second AI video at $0.07 a second, OpenAI closes $11 billion at a $1.25 trillion valuation, and SMEE shows new lithography machines.

By H.AI Tools Team3 stories
Products in this issue: Gemini, Google Veo, Sora, Runway, ChatGPT, Claude, Perplexity
In this issue

The short version

  • Gemini Omni 1.2 Flash generates video clips of 3 to 15 seconds at $0.07 a second for 720p. You can draft at low resolution and upscale only the take you keep.
  • OpenAI reportedly closed the first $11 billion of a new round at a $1.25 trillion valuation, below its earlier target. OpenAI has not announced it.
  • Shanghai Micro Electronics showed models of its SSC800 and SSB800 lithography machines, which serve mature chip processes, not leading-edge ones.
  • Heavy AI funding keeps prices and plans changing quickly, so prefer monthly commitments while products are still moving.

The November 9 issue starts with a tool you can use immediately: Google's cheaper, longer AI video model. It then looks at the money and machines behind AI, with a large funding close at OpenAI and new Chinese chipmaking equipment on show in Shanghai. Each section separates what was confirmed from what was reported.

01

Gemini Omni 1.2 Flash: 15-second AI video with draft-then-upscale

What Google released

Google DeepMind released Gemini Omni 1.2 Flash as a faster, cheaper multimodal video model. Native clip length rises from 10 to 15 seconds, and you can choose any length from 3 to 15 seconds. You can generate a quick draft at 360p or 720p, then enlarge the one you approve to 1080p or 4K. Google says characters, objects and scenes stay more consistent within a clip, and you can revise a clip over several turns of conversation instead of starting again.

Pricing for 720p output through the API is $0.07 per second, about 30% below the previous version, so a 15-second clip costs about $1.05 before any upscaling. Google Workspace users can also make 15-second clips in Google Vids at no extra cost.

A draft-first video workflow

The draft-then-upscale design changes how to budget for AI video. Most spending goes on takes you discard, so generate drafts at the lowest resolution that lets you judge the shot. Spend on high resolution only once a take is approved.

Try a 15-second product clip. Write one paragraph covering the subject, setting, camera movement and the action at second 0, second 7 and second 15. Generate three 360p drafts in Gemini, pick the best, then refine it in conversation: "keep the camera move, make the background lighter". Upscale only that version. Record the total cost and number of turns. Repeat the same brief in another generator such as Google Veo, Sora or Runway to see which gives a usable clip for the least effort. Check commercial-use terms and any rights in reference images before publishing. See our Sora vs Runway comparison and the Video Generation category for more.

Get HAI Daily in your inbox

The day's AI news, what it changes, and which tools to try.

02

OpenAI's $11 billion close and the cost of building AI

What was reported

OpenAI closed $11 billion of new funding as the first part of a round valuing it at $1.25 trillion, according to people familiar with the matter cited by Reuters and Bloomberg. SoftBank and MGX said they participated. The round was originally pitched at up to $30 billion at a higher valuation. Reports now expect it to total around $20 billion. OpenAI has not announced the financing.

Anthropic is also preparing a public listing. It launched its IPO roadshow the following day, covered in our November 12 issue.

What heavy funding means for your subscriptions

For users, the visible effect of this much capital is constant change in products and prices. Providers launch new tiers, cut some API prices, add usage limits to others and bundle features into higher plans. That is good for capability but makes long commitments risky.

A few habits help. Prefer monthly plans while a product is changing quickly, and move to annual billing only for tools your team already uses every day. Once a quarter, compare what you pay for ChatGPT, Claude or Gemini with what you actually used, using the usage pages each provider offers. When you read funding news, check whether a figure is announced by the company or reported by sources. Answer engines like Perplexity make it quicker to find the original report. This is general purchasing guidance, not investment advice.

03

SMEE's SSC800 and SSB800: China's lithography for mature chips

What was shown

At the China International Import Expo in Shanghai, Shanghai Micro Electronics Equipment (SMEE, 上海微电子) displayed models of two front-end dry scanning lithography machines. The SSB800 series uses i-line light and targets processes at 280 nanometres and above, used for power semiconductors, analog chips and similar components. The SSC800 series uses KrF light and targets 110 nanometres and above, serving high-density storage and AI high-performance computing. SMEE says key subsystems are developed in China. These include high-precision optics, illumination and a fast dual-stage platform that moves the wafer.

These are mature-process machines. They are not the extreme-ultraviolet (EUV) systems used to make the most advanced AI processors. Media reports added production figures, such as cumulative wafer output and uptime rates, but we could not find them in an official SMEE statement and do not repeat them here.

Reading chip-equipment news with care

AI chips depend on many manufacturing steps, and a large share of the components in an AI server, such as power management and memory interfaces, is made on mature processes. Domestic supply of mature-node lithography therefore matters for China's electronics supply chain even though it does not close the gap at the leading edge.

When you follow chip news, note three things for each announcement: the process node a machine serves, whether it is a model on display or a shipping product, and which figures come from the company itself. Search tools in the Search Engines category can help you find the manufacturer's own statement. Company-research tools in Finance AI help track the businesses involved.

Not sure which tool fits your task?Browse the H.AI Tools directory by category, then compare a shortlist.

What to try next

Try the new video model, keep commitments flexible

If you make video, test Gemini Omni 1.2 Flash with a draft-first workflow and compare cost per usable clip with your current tool. Review your AI subscriptions against actual use, and stay on monthly plans until a tool is part of your daily work. Read chip and funding headlines for what is confirmed, not only for the largest number.

Explore Video Generation and Chatbots, and see our AI video generator guide for a broader comparison.

How we put this issue together

The H.AI Tools Team checks each story against the official sources linked in the text and removes claims we could not confirm. Product links lead to H.AI Tools directory profiles. Unless an issue says otherwise, product notes describe vendor documentation, not our own benchmark tests.

Go deeper