FrameQuery uses AI to analyze, index, and make your video content searchable with natural language.
Find anything across your video library
A003_C008_0342.R3D
"Lena (Director) at 03:22, 11:45..."
B002_0147_1012.braw
"...now for the product reveal..."
MVI_2847.mp4
"detected at 00:14, 02:38..."
Drop in any format. FrameQuery indexes everything, then gets out of your way.
Connect your sources. FrameQuery processes any format and builds a local index.
| File | Source | Duration | Size | Scenes | Speakers | Status |
|---|---|---|---|---|---|---|
R3DA001_C012_interview.R3D | /Volumes/RED_RAW_2025 | 01:14:32 | 18.4 GB | 24 | 2 | Indexed |
BRAWB_Roll_Drone_4K.braw | /Volumes/BMPCC6K | 00:22:15 | 6.2 GB | 8 | 0 | Indexed |
MP4Conference_2025.mp4 | ~/Documents/Interviews | 00:32:15 | 245 MB | 11 | 4 | Indexed |
MP4Podcast_EP47_Recording.mp4 | ~/Documents/Interviews | 00:58:02 | 312 MB | 3 | 2 | Indexed |
R3DA002_C005_cutaway.R3D | /Volumes/RED_RAW_2025 | 00:38:47 | 9.1 GB | 14 | 1 | Indexed |
MOVOffice_Tour_ProRes.mov | Dropbox/Marketing | 00:14:20 | 4.8 GB | 6 | 1 | Indexed |
BRAWBMPCC_Closeups_Day2.braw | /Volumes/BMPCC6K | 00:45:55 | 11.3 GB | 19 | 0 | Indexed |
MOVproduct_demo_final_v3.mov | Dropbox/Marketing | 00:08:44 | 1.1 GB | · | · | Processing |
R3Dbehind_the_scenes_raw.R3D | /Volumes/RED_RAW_2025 | 02:05:10 | 31.7 GB | · | · | Queued |
Every indexed video in one place. See scenes, speakers, and processing status at a glance.
~/Documents/Interviews
/Volumes/RED_RAW_2025
Dropbox / Marketing
/Volumes/BMPCC6K
Connect folders, drives, or cloud storage. New files are picked up automatically.
Find any moment across your entire library by words, faces, objects, or scenes.
Conference_2025.mp4
"...we need to focus on our quarterly goals and improve our marketing strategy..."
meeting.mp4
Scene: Marketing team around conference table with laptops and whiteboards
library.mp4
Scene: Team meeting discussing quarterly goals in a casual breakout setting
A001_C012_interview.R3D
Sarah (CEO) detected at 04:10, 21:44, 38:02 and 5 more timestamps.
Search across transcripts, detected objects, scene descriptions, and people with natural language. Filter by duration, date, colour, source, and more.
Jump straight to any clip, copy its file path, or export a precise cut to your editor.
CEO delivers opening keynote to packed conference hall
Product team walks through Q2 roadmap on screen
Marketing team reviews quarterly goals and growth targets
Customer panel shares feedback on the latest release
Teams break into groups for hands-on feedback session
Executive sits down for a focused one-on-one interview
Every scene described and tagged with objects, colours, and timestamps.
Our company has grown by 35% year-over-year and expanded into three new markets.
I'd like to share updates about our Q2 product roadmap.
We need to focus on quarterly goals. This quarter is critical.
Our Q2 targets are ambitious but achievable if we maintain momentum.
AI transcription with named speaker labels and click-to-seek timestamps.
Map names to faces and voices. Find every clip a person appears in.
Format
Export as FCPXML, Premiere XML, EDL, or JSON with full metadata - straight into your editor.
We reccommend starting with Pro to process your backlock of videos.
Planned pricing - subject to change before launch.
Search your own index. Always free
Once a video is processed, the index is saved locally on your machine. Searching it is always free, even offline. See supported formats →
Everything you need to know about FrameQuery.
FrameQuery is ideal for video editors, content creators, marketing teams, and anyone with a large video archive.
Yes. Searching your own local index is free. All search features are available on the free tier.
Yes. After processing, the metadata and search index are saved locally. You can keep it, back it up and search it any time.
An index is a saved set of searchable notes about a video. It includes what was said (with timestamps), what appears on screen, and scene summaries. FrameQuery stores it locally so you can search later without paying again.
FrameQuery supports many video and audio codecs and formats. See our compatibility page for the full list and per-format details.
Yes - export your selections as FCPXML, Premiere XML, EDL, or LosslessCut JSON to drop directly into your NLE. A public API is also on the roadmap.
We only charge for things that directly cost us money. We use cloud compute for video processing and for hosting shared indexes. Since searching your own index is local, it's totally free and will still work even if we disappeared.
Absolutely. We use industry-standard encryption and security practices. Your video proxies are deleted immediately after processing and your data is stored securely and never shared with third parties.
A general metric is around 5 minutes per hour of video once uploaded. Transcriptions can take longer depending on video length and queue depth.
FrameQuery uses a number of techniques to minimise the amount of data processed without compromising on search quality. We pass these cost savings down to you.