NoodleTomato is an AI video generation tool for faceless YouTube videos. Its core idea is: users only need to describe the video topic in one sentence, and the agent will automatically complete the entire video production process—researching materials, writing the script, voiceover, shooting, and editing, ultimately outputting a complete video. Video duration ranges from 10 seconds to 2 hours and can cover any topic.
In terms of main capabilities: the agent works in stages and shows the user progress and results at each stage, allowing the user to review and guide. The tool has a built-in research step: first gather materials, then write a brief, and finally write the script. The official description says the narration content will not be fabricated out of thin air. The video visuals and audio are generated in one pass by the self-developed model NoodleTomato Motion, at about $0.03/second. In terms of style, the page offers fifteen styles and also supports users describing their own style. Output formats support widescreen, vertical, or square, and different formats can be generated in the same run. The voice library has tens of thousands of narration and dialogue voices; each video can choose one or more voices, or none. The content memory feature can save narration, appearance, characters, and information from each previous episode, so the next production can continue from the previous progress, keeping the same characters (same face, clothing, voice) consistent within the same series. Users can also make modifications according to their own requirements, and the agent only changes the specified parts. In addition, the agent can also work with real footage, rather than requiring AI-generated visuals.
In terms of usage, the user enters a one-line description (brief), and the agent takes over the entire production process and presents it in stages. Users can also skip the agent, write their own prompt, and directly use NoodleTomato Motion to generate clips with audio and visuals from text or reference images. It also supports connecting your own agent via MCP (when using your own agent, inference costs are paid on your own client, and NoodleTomato only charges credits for media generation), as well as programmatic calls via API for text-to-video, image-to-video, reference-image generation, etc. The API provides quotation, credit reservation, status query, task cancellation, asset download, revocable limited-permission keys, and other features.
Applicable scenarios include: production of documentaries, product ads, short dramas, courses, and various other videos, especially suitable for weekly-updated YouTube channels or series content. The official page mentions that the finished video is an asset owned by the user, and the most common use is publishing to YouTube, but it is not the only use.
Limitations and notes: Use of the studio, agent, MCP, and API requires a subscription to anyvalid plan. Credits are billed at 100 credits = $1. Credits obtained from a plan or top-up expire after 2 calendar months from the date they are issued; renewing or topping up adds a new batch of credits and does not extend the old credits. Each task is quoted before running, and the actual cost will not exceed the quote; regeneration or modification counts as a new generation. Videos are billed by the generation seconds of the selected options, images are billed per image, voice and music are billed per use, and NoodleTomato-hosted inference is billed per run. Specific prices are, for example: NoodleTomato Motion landscape is $0.0335 per second, Motion HD starts at $0.0625, reference images start at $0.0041; narration is about $0.34 per minute of speech ($0.4888 per 1,000 characters); music background tracks are $0.069 each; agent inference is about $0.0007 per step. Plans range from $9 per month (900 credits) to $5,000 (500,000 credits).
