Process recorded webinars
Connect
SPIRITT
VLM RunVLM Run extraction.
Any media, structured.
SPIRITT uses VLM Run to read photos, documents, recordings and video like a trained analyst. It extracts structured JSON with the right schema, runs your multimodal agents and pushes every result into the systems that act on it.
What you can do with VLM Run.
Structure from any media
Images, documents, video and audio become JSON via prebuilt domains, custom schemas or skills.
Multimodal agents on demand
Runs your VLM Run agents over files and collects results and artifacts.
Reusable skills
Packages prompts and schemas into skills so every run stays consistent.
Works with your other apps.
Get started in three steps.
01
Open your workspace
Start a SPIRITT workspace for your media pipeline and the systems that use its output.
02
Connect VLM Run
Connect VLM Run with your API key so SPIRITT can upload files and run extractions.
03
Delegate an outcome
Share the media and the fields you need. SPIRITT extracts, checks and delivers the data.
Hand over a mission.
Choose a mission. Make it yours.
Put VLM Run to work.
Every file, turned into data
Questions
VLM Run, with SPIRITT.
01What can SPIRITT do with VLM Run?+
It uploads files, runs structured extraction on images, documents, video and audio, executes agents, creates and finds skills and tracks runs and artifacts.
02How do I connect VLM Run?+
Open a SPIRITT workspace, connect VLM Run with your API key and share the media you want processed.
03Which media types work?+
Images, documents, video and audio.
04Can I use my own schema?+
Yes. SPIRITT uses a prebuilt domain, your custom JSON schema or a saved skill for each extraction.
05How are long jobs handled?+
Document, video and audio runs are asynchronous. SPIRITT follows each run or receives a callback, then continues the workflow.
06Can it run my VLM Run agents?+
Yes. SPIRITT lists your agents, executes them on multimodal inputs and collects their results and artifacts.