Describe what you want. InferenceVideo writes the script, generates the imagery, narrates it, and renders a finished, chaptered video — in minutes.
InferenceVideo orchestrates the best of each — you never touch the plumbing.
Turns your brief into a structured, chaptered script and a slide-by-slide deck.
Generates bespoke, cinematic imagery for every scene — no stock, no clip-art.
Voices the script naturally, in the tone and pace you choose.
Records it all to the narration and delivers an .mp4 plus a self-contained, chaptered web player.
Type the topic, paste an outline, or drop in a document. Pick a length, a tone, and a narration voice.
Claude proposes a chaptered script and deck. Edit any beat inline — you stay in control.
Imagery and narration render beat-by-beat. Watch it come together.
Preview in the browser, then take your .mp4 and the interactive, chaptered player. It's yours.
A product launch, a course, an invite, a brand film — the same studio, wildly different results. Every reel below was made by InferenceVideo. Soon you'll share and browse films on Inference TV.
11 films across 9 categories — filter by type, play any with sound, and hear Pablo critique each one.
InferenceVideo is the first nosource product — released free of charge, permanently. You connect your own Anthropic, OpenAI, and ElevenLabs keys, so your usage is entirely yours and the studio stays free.
Zero cost to use, forever. That's the nosource promise — software built with AI, given freely.
Paste your provider keys once. Encrypted at rest, used only for your renders, never shown again — or keep them session-only.
Your briefs, your videos, your accounts. Sign in with Google; everything you make is downloadable and portable.
Tell us what you'd make. We read everything.