Native video understanding.One API.
Scroll to explore
In-Video AI + Agents
Every video searchable, clip-ready, understood.
Understand, summarize, segment, and search video — switch capabilities and click results to seek.
Know what happens in every frame.
Read scenes, actions, speech, and on-screen text together, and get structured JSON describing what the footage actually shows.
Use cases
Built for real footage.
Robot demonstrations, unlabeled archives, long-form media, sensitive records. Same call, same time-anchored output.
Media & Entertainment
Work through an archive instead of a timeline. Long-form footage comes back as sampled frames, scored moments, and trimmed segments, so editors pull an edit-ready cut without scrubbing the master file.
- 0:01sampled 12 keyframes from master
- 0:03segment scored for highlight reel
- 0:04trimmed cut exported with watermark
Pricing calculator
Video understanding by the hour.
Pick a workload, footage volume, and model. See cost savings versus vendors that bill every second of the timeline.
Footage volume / month
Gateway billed tokens (in / out)
2.4B - 3.5B / 300M
Gemini 3.5 Flash Lite billed tokens (in / out)
10.4B / 600M
Total costwhat the gateway bills for this workload
$335 - $2K
Total cost savingsvs the cheapest frontier model
$2.7K - $4.3K
Video / month
10Khours
Cost comparison (log scale)
Cheapest - most expensive
For Enterprises
The new visual intelligence layer for your enterprise.
Deploy securely inside your VPC or private cloud — bring video understanding to your infrastructure without shipping footage to a third-party SaaS. SOC 2 Type II and HIPAA-ready.


Video AI FAQs
Caption and summarize clips, transcribe speech with timeline anchors, extract highlights, and search long-form libraries — all through the same OpenAI-compatible API you use for documents and images.
