跳到正文
DEV Community· T.M. Gunderson·· 2 天前AI 評分63

Gemini Omni 1.2 Flash 影片生成每秒 0.07 美元,小型企業市場推廣預算如何改變

The $1 Video: What 7-Cent-Per-Second AI Generation Does to a Small-Business Marketing Budget

AI 導讀

Google 將 Gemini Omni 1.2 Flash 的 720p 影片生成費降至每秒約 0.07 美元,15 秒影片約需 1 美元。按文中的月度安排,4 條 15 秒影片生成費約 4.20 美元,圖像及 GIF 生成另需數美元。作者指出,策略、品牌語調、商戶實拍和問責等環節,仍是聘用自由工作者或代理公司的價值所在。

正文

Google made Gemini Omni 1.2 Flash generally available on November 5, 2026, and the price cut that came with it is the kind of move that rewrites a small business's content math. The model now costs about $0.07 per second of 720p video through the Gemini API — down 30% from the $0.10/second it launched at in June. A 15-second clip costs roughly a dollar. Two years ago, that same clip would have cost more than most owners spend on AI tools in an entire month.

The natural question is: what does this actually mean for your marketing budget?

The Monthly Cost Math

A realistic small-business content calendar might include 4 short videos, 8 GIFs, and 20 still images per month. The video piece, at 15 seconds each on Gemini Omni 1.2 Flash, costs about $4.20 total — less than a latte. Image and GIF generation from the same generation of tools adds only a few more dollars.

The entire production run is roughly the price of one line item on a freelance invoice.

That is not a typo, and it is not theoretical. The pricing is live on Google's Gemini API right now: 5,792 video tokens per second billed at $12 per million output tokens, which works out to approximately $0.07 per second at 720p. Other resolutions run about $0.023/second at 360p, $0.104/second at 1080p, and $0.209/second at 4K.

OpenAI's Images 2.6, released November 4, points the same direction: image generation is now ~50% faster than its predecessor, with GIF creation from prompts or sketches added to the mix. Both releases share one signal — the cost of generating a marketing asset is heading toward rounding error.

What Stays Expensive (The Honest List)

If production is now cheap, what are you actually paying for when you hire a freelancer or agency? Here is the honest answer:

Strategy and brand voice. AI generates the asset. It does not decide which asset to make, what story it tells, or whether it matches the voice your customers already trust. That judgment is still human work.

Footage of your actual business. Omni Flash generates generic scenes. It does not walk your shop floor, film your staff at work, or capture the specific thing that differentiates you from the competitor three blocks away. Stock-looking video is cheap. Your actual story still requires a camera or a budget to hire someone with one.

Accountability. When an AI-generated video performs badly, the tool does not explain why. An agency or freelancer who knows your business, your audience, and your past campaigns can. That relationship is what the invoice is actually for.

Editing judgment. Gemini Omni 1.2 Flash supports multi-turn editing — you can swap backgrounds, change lighting, or modify scenes across several chat turns. But users report character consistency begins to drift after roughly three to five turns, and safety filters can block prompts without refunding the generation cost. The tool is fast. Deciding whether the output is good enough to publish is still a human call.

Three Questions Before You Scale

Production-cost collapse creates a specific temptation: making more content just because you can. Before expanding your calendar, ask:

  1. Does this piece serve a customer decision? Content that answers a real question at a real step in the buying process outperforms content that exists because generation is now cheap.

  2. Can a customer tell it is generic? AI-generated video is getting harder to spot, but it still tends toward the visually plausible and narratively bland. If your differentiator is personal service, local knowledge, or craft quality, generic video can actually undermine the message.

  3. Do you have a system to test what works? Cheap production without a feedback loop becomes digital hoarding. Know how you will measure whether the 15-second clip drove a call, a booking, or a sale before you generate it.

What This Means for the Agency Relationship

If a freelancer used to charge you $500 for four short videos and the raw generation now costs $4, the value proposition has shifted. The fair question to ask any agency or creator you work with is: what part of your fee is production, and what part is judgment, strategy, and accountability?

The honest ones will separate the two. The ones who do not are selling you a production service in a market where production just became nearly free.

Honest Boundaries

  • Prices change. Google's $0.07/second rate is current as of November 2026. Rates have moved before and will move again.
  • Quality varies by use case. Omni Flash excels at conversational editing and consistency across short clips. Fast action, complex physics, and quick script changes are still listed as active limitations by Google.
  • The 15-second ceiling per generation means longer-form content requires stitching, which adds time and potential quality drops at the seams.
  • Content filters are real and occasionally overreach. Google users have reported original fictional characters being blocked by "prominent people" filters, with no refund for the failed generation. Budget a small buffer for retries.

The cost collapse is genuine. The opportunity is real. But the businesses that benefit will not be the ones that generate the most content — they will be the ones that generate the right content, with a human still making the call on what "right" means.

If your business already pays for AI tools and marketing support but you are not sure which pieces are actually earning their keep, the AI Clarity Sprint ($1,495 CAD, fixed fee, two weeks) maps every AI touchpoint you already have, then hands you a ranked plan for what to do first.

來源:DEV Community · dev.to