LONG AI
Home › Blog › Omni 1.2 Ships: Video Prices Drop 30% Fr

Omni 1.2 Ships: Video Prices Drop 30% From Nov 6

Category:资讯 · 6 min read · 2026-11-06

In the early hours of Nov 6, Google put the video generation model Gemini Omni 1.2 Flash on the Gemini API and Google Cloud, and cut the price by a third. 720p video went from $0.10 to $0.07 per second; a 15-second clip drops from roughly ¥10 to ¥7 in RMB. For short-video, e-commerce, and ad work, this is the most substantive video-API price cut of the year.

Gemini Omni 1.2 Flash 30% price cut: 720p from 0.10 to 0.07 USD/second, a 15s clip about 7 yuan, announced Nov 5, not on our plaza
Fig 1 - Left: a 15-second clip's official cost drops from ~10 to ~7 yuan; middle: the four things that changed; right: our plaza test on Nov 6. Data from Google's blog and our API, self-made.

1. What the official side said

Source: Google's official blog Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips, published 2026-11-05 17:00 UTC (01:00 Beijing time Nov 6), by Anish Nangia (Google DeepMind PM); the publish-time field is verified.

Two key sentences, quoted directly: "We have dropped the price 30% from Gemini Omni 1.1 making it available at $0.07 per second for 720p video." "This model is generally available on the Gemini API and Google Cloud for your production needs." In plain terms: 30% cheaper than last gen, $0.07/second at 720p, and already generally available. A preview can change anytime; only GA is safe for production.

2. What changed: more than price

  1. Clip length from 10 to 15 seconds, and you can specify 3 to 15 seconds in 1-second steps. Before, a full shot often needed two clips stitched together; now it's one take.
  2. Multi-turn conversational editing: swap the product, change the background, adjust lighting/weather, change aspect ratio - all as edits on the same clip, no full regeneration. This is the real money-saver.
  3. Reference images behave better: people, products, style references, even up to 3-second motion clips and audio prompts can be fed in, with less drift from first to last frame.
  4. Multi-person dialogue is more stable: lip-sync holds, and you can specify who speaks and with which voice.

3. The math: a 15-second clip drops from 10 to 7 yuan

ItemOmni 1.1 (old)Omni 1.2 (new)
720p per second0.10 USD (back-calculated)0.07 USD (verbatim)
A 15-second clip1.50 USD ≈ ¥10.041.05 USD ≈ ¥7.03
Per-clip differencesaves ¥3.01, a 30% cut

The math, so you can re-verify:

4. Can you buy it here: not today

On the morning of 2026-11-06 we pulled our /api/v1/model-plaza API and cross-checked the page:

The conclusion is blunt: to use Omni 1.2, the only route today is Google's official channel (Gemini API or Google Cloud). No seller here lists it - can't search it, can't call it.

Plaza search for omni-1.2 shows no matching models, top shows 185 models and 924 listings, confirming Omni 1.2 Flash not yet integrated
Fig 2 - Our plaza on 2026-11-06: "omni-1.2" returns zero, top-of-page 185 models, 924 listings.

One more caution: plaza video models quote per generation, while official Omni 1.2 bills per second; how many seconds one generation equals varies by vendor, so the two can't be subtracted directly to say who's cheaper - align the seconds first.

5. What it means for ordinary users

  1. Already on Omni 1.1: same clip is 30% cheaper and can stretch to 15s - pure win.
  2. On the veo family: unaffected. veo didn't change price and no migration was requested.
  3. Occasional video generation: three yuan per clip isn't huge, but 15s finishes the shot in one pass, fewer reworks - what you save is time.
  4. Text-only or coding: this one's irrelevant to you, skip it.

6. Our advice: five you can act on

  1. Draft low-res, then upscale: the official side explicitly says draft at 360p or 720p, pick what you like, then upscale to 1080p or 4K. Don't start at max - one rework is one clip's cost.
  2. Edit with "continue editing," don't regenerate: background, lighting, aspect-ratio changes all go through multi-turn editing - the official side names this the main money-saver.
  3. Budget per clip: a 15s clip is about ¥7. Count how many you need this week and multiply - far more intuitive than watching unit price.
  4. To order here, use veo-3.1 for now: all three plaza versions are callable now, billed per generation with multiple sellers bidding. Once Omni 1.2 has supply, it'll appear on the plaza automatically.
  5. Don't rush migrating old projects: the official side only announced a price cut, no Omni 1.1 retirement timeline. Leave running work alone.

Scope note

Official info from the Google blog verbatim (2026-11-05 17:00 UTC, publish time verified), price and capability copied as written; "Omni 1.1 at $0.10/s" is back-calculated from the "30% off" claim, marked in the body. RMB conversion at our uniform rate 6.695. Plaza in-stock status, model count, and seller count from our /api/v1/model-plaza API and page, fetched 2026-11-06 09:45. This article doesn't judge quality; plaza price moves with supply, and the final price follows the model plaza page.

相关文章

GPT-6.2 Cuts Prices 20%: What the AI Price War Means

GPT-6.2 cuts list prices 20%, yet platform discounts run deeper: gpt-6-astra trades at under 1% of list. What this week's AI price war means for buyer

What Is Prompt Caching? How Cache Tokens Cut API Bills

Prompt caching stores a repeated prompt prefix and bills cache reads at 5-12% of normal input. Real prices, hit conditions, and how to cut your API bi

All 12 Claude Models Compared: Price and Best Use

Twelve Claude versions are on sale. The newest opus-5-5 costs $8.91 a month; the oldest opus-4-6 costs $49.50. Full table and a pick for each workload

Which AI Model Writes Code Best? Speed and Price Tested

gpt-5.6-terra and deepseek-v4.1-flash win on price and reliability; gemini-3.8-flash is fastest. Effective prices, measured speed, and a three-tier pi

One API Key, Every Model: How to Integrate a Multi-Model Marketplace

Point your base URL at the platform and one key covers all 138 on-sale models across 51 vendors, in both OpenAI and Anthropic request formats. Copy-pa