I want to create about three short videos and a few images per month for social media. Assuming such personal use, I have compared the currently available AI services from a cost-performance perspective.
First, as a major premise, Sora (OpenAI) is no longer available. The web version and app were discontinued on April 26, 2026, and the API on September 24 of the same year, so it is already off the table as a means of video generation, including via ChatGPT. There has been no announcement of a successor model.
Price and Feature Comparison Table (as of November 2026)
*1 Meta has announced video generation as “Muse Video,” but as of November 2026, it is in the early preview stage of a closed beta and has not been released to the public. *2 ChatGPT Plus has had its video generation feature removed following the discontinuation of Sora, and currently only supports image generation.
Note that “Muse,” which began service in Japan on October 23, 2026, is a personal AI agent that handles tasks like sending emails and making reservations; it is a different product from “Muse Image/Muse Video,” which is also made by Meta but performs image and video generation. Please be careful, as the similar names make them easy to confuse.
Conclusion 1: “Google AI Plus” is the best choice for 3 videos + images per month
If you want to create about three short videos and some images together per month, the conclusion is that **Google AI Plus (725 yen/month)** has the best cost-performance. The reason is simple: it is the only plan that allows you to use both image generation (Nano Banana 2.1) and video generation (Gemini Omni) for a price under 1,000 yen. ChatGPT Plus costs 3,000 yen per month and cannot create videos at all right now.
If images are enough and you don’t need videos yet, then operating for 0 yen with the free version of Gemini or Meta AI’s Muse Image is perfectly possible. I think the most cost-effective order is to try those free options first, and then switch to Google AI Plus when you truly need video.
Conclusion 2: “Muse” is also worth trying for delegating tedious tasks
Up to this point, we have been talking about image and video creation, but “Muse,” which arrived in Japan on October 23, 2026, is a different type of personal AI agent. Since you can entrust it with tasks like sending emails, making reservations, and shopping, it is for people who want to “save working time” rather than for content creation itself. Basic features can be used for free.
If you use a referral code and enter it in the settings screen within 48 hours of registration, we will both receive 1 billion Muse tokens. If you are interested, please use this link.
Code: U7T6XY 🔗 https://muse.ai/join
*This is a benefit granted via a referral code.
Points to note
The pricing, free tiers, and supported features of AI services are areas that change frequently. This article is based on information researched as of November 2026, but please check the latest prices and specifications on each service’s official website before using them.
This article is based on independently researched information as of November 2026. Prices and specifications are subject to change.
