Chat2Video Cloud guide

Chat2Video Cloud makes footage queryable by what was said. Upload a video together with its .srt or .vtt transcript and every subtitle line becomes searchable evidence, tied to the moment it was spoken.

Step-by-step

1. Open Chat2Video Cloud and sign in.
2. Create a collection (e.g. "Site diary — October").
3. Upload footage — MP4, MOV, WebM or audio, plus a matching .srt/.vtt transcript with the same file name. When the deployment has an AI Gateway key, audio is transcribed automatically; otherwise the transcript is required for searchable speech. Hosted deployments cap a single upload at about 4.5 MB.
4. Wait for ingest — transcript lines are chunked with their timestamps. Progress is shown per file.
5. Ask:

"When did the concrete truck arrive on the 14th?" "What did the foreman say about the pour sequence?"

Answers cite timestamps down to the second — click to jump to the moment.

Transcript search

Search what was said across every clip in a collection. Matches come back as moments, not whole files, so you can jump straight to the relevant exchange. Upload the transcript beside its recording to keep citations attached to the right video.

Practical notes

• Ingest is fast because it indexes text, not frames.
• One collection per site/month keeps timestamp answers easy to read.
• Mixed collections work: PDFs, Office files and transcripts can be queried together, with each citation pointing at its own source.
• Frame OCR and object tracking are not part of the in-app engine; upload a transcript for speech-based questions.

Free tier

20 minutes of footage/month, 1 collection. See pricing.