Integrations
The homepage credits Supabase, Vercel, JigsawStack, Hasura and Fly as the infrastructure behind the engine, and links out to a public GitHub repository and the author's X profile.
AI Video Search Engine
Ask a question, get answers pulled from inside the videos in the index.
Beginner
Own AI Video Search Engine? Verify ownership to get a dofollow link
What is AI Video Search Engine?
AI Video Search Engine is an AI search tool. A question answering search engine over an indexed video library. Ask in natural language, get related answers from the footage, then chat with any indexed video. The homepage reports 17,272 videos and 277,758 minutes indexed, with source code on GitHub.
AI Video Search Engine is a search layer built on top of a video library rather than on web pages. Instead of matching keywords against titles and descriptions, it indexes the spoken and visual content of each video so a question in plain language returns the relevant passage, the surrounding context, and the video it came from. The homepage reports a current index of 17,272 videos covering 277,758 minutes of footage, which gives a sense of both the scale and the freshness ceiling of the dataset. The workflow is deliberately simple. A user types a question, the engine returns related answers drawn from indexed videos, and a chat mode lets the user keep asking follow up questions about a specific video. That makes it useful for researchers who need to verify a claim without rewatching an hour of footage, for learners working through long lectures, and for anyone who has a large archive of recorded talks and wants to make it queryable. The homepage groups indexed material into broad topics such as human brain, Supabase, startups, AI image generation and the future of startups, which suggests the current corpus leans toward technical talks and founder interviews. Under the hood the project is assembled from hosted services rather than a single proprietary stack. Supabase, Vercel, JigsawStack, Hasura and Fly are listed as the supporting infrastructure, and the source code is public on GitHub. That transparency is unusual for a search product and is the main reason to consider it if self hosting or forking is on the table. As an alternative to general web search or to YouTube's built in search, the value depends entirely on whether your topic is already in the index. It is a narrow, transparent, developer friendly experiment in conversational video retrieval, not a universal video search engine, and buyers should treat the index size as the first thing to check.
How do you use AI Video Search Engine?
1Open the engine and review the index stats
Load the homepage and check the counter showing total videos and minutes indexed. That number tells you how much material the engine currently has to search.2Pick a topic or type your own question
Use the topic tiles (human brain, Supabase, startups and others) as a starting point, or enter a natural language question in your own words.3Read the returned answers and source videos
Review the related answers the engine surfaces and open the underlying video to confirm the surrounding context before citing anything.4Start a chat on a single video
Once you have found a relevant video, switch to chat mode to ask follow up questions about that specific recording rather than the whole index.5Sign in if prompted
Some actions require an account. Sign in when the interface asks, then return to your search or chat session.6Inspect or self host from the repository
The homepage links to the GitHub repository and to the author on X. Use those links if you want to read the code or run your own instance.
Pros and cons
Pros
- Answers come from inside the video content, not just titles or descriptionsAI
- Chat mode supports follow up questions on a single videoAI
- Index size and total minutes are published openly on the homepageAI
- Source code is public on GitHub, which makes the approach inspectableAI
- No account barrier for browsing the existing indexAI
Cons
- The index covers only 17,272 videos, so niche topics may return nothing usefulAI
- Topic coverage looks skewed toward technical talks and startup interviewsAI
- No documented pricing, support channel or API terms on the homepageAI
- Search quality cannot be judged without knowing how transcripts and visuals are chunkedAI
- Sign in is required for some actions, but the homepage does not say which onesAI
How much does AI Video Search Engine cost?
Features
Natural language question answering over an indexed video corpus, related answer retrieval with source videos, chat mode for follow up questions on a single video, topic grouping across indexed material (human brain, Supabase, startups, AI image generation, future of startups), and a public counter reporting 17,272 videos and 277,758 minutes indexed.
Frequently asked questions about AI Video Search Engine
How do you use AI Video Search Engine?
The walkthrough on this page covers 6 steps: 1. Open the engine and review the index stats 2. Pick a topic or type your own question 3. Read the returned answers and source videos 4. Start a chat on a single video 5. Sign in if prompted 6. Inspect or self host from the repository.What platforms does AI Video Search Engine support?
AI Video Search Engine is available on Web App.What does AI Video Search Engine integrate with?
The homepage credits Supabase, Vercel, JigsawStack, Hasura and Fly as the infrastructure behind the engine, and links out to a public GitHub repository and the author's X profile.What are the limitations of AI Video Search Engine?
The index covers only 17,272 videos, so niche topics may return nothing useful. Topic coverage looks skewed toward technical talks and startup interviews. No documented pricing, support channel or API terms on the homepage.
Status
StatusActive
Views0
Outbound clicks0
Added10/21/2026
Platforms
Web App
Pricing
Free
