Selendis introduces Archon, a video model that turns raw footage into finished edits
Overview
Selendis has introduced Archon, a video foundation model that turns raw footage into a finished edit without task-specific training, according to a post by Rohan Paul on X. The post says Archon learns each creator's cutting style from accept and reject decisions in Selendis's multiplayer editor, from the creator's past videos, and from reference videos the creator cites. Selendis's own announcement says early access is offered to users who comment "SELENDIS", with three months free. The post describes Archon's taste model as a per-user prior that keeps updating, but it does not report measured editing quality or user results.
Written by AI from the articles below · updated Nov 3, 12:29 PM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
Rohan Paul@rohanpaul_aiXSelendis's Archon video model learns creator taste from editing behaviorAISelendis has introduced Archon, a video foundation model that turns raw footage into a finished edit zero-shot and keeps learning each creator's cutting style. The taste signal comes from accept and reject decisions in Selendis's multiplayer editor, a creator's past video catalogue, and reference videos the creator cites. Per the post, Archon treats taste as a learnable per-user prior.
Heat Trend
Not enough continuous observations to show a trend yet.