Skip to main content

TL;DR: The most robust Jumper workflow

  1. Download and select the Ultra Accurate analysis model in Jumper’s Settings. See the visual model reference for compatibility and requirements.
  2. Analyze the footage for Visual + Speech + Face. Select the footage in the Media tab, then run visual analysis, speech analysis, and face analysis.
  3. Give names to the detected faces and transcript speakers. Name faces in the People tab and speakers in the transcripts.
  4. Find the shots you need directly in your NLE. Use Jumper inside Adobe Premiere Pro, Final Cut Pro, DaVinci Resolve, or Avid Media Composer to search and add footage to your edit.
  5. Or use Claude, Codex, or ChatGPT for agentic editing. Ask the agent to start a Summary Analysis and tell it what timeline, analysis, deliverable, or task you actually want. Summary Analysis creation and the final deliverable can be part of the same prompt, so you can kick off work that runs for a long time without needing any further input.
jumper_in_premiere_smile.avif

Integrated with Editing Software

Works inside Adobe Premiere Pro, Final Cut Pro, DaVinci Resolve, and Avid Media Composer

Local and Offline

Everything runs locally and offline on your device. No cloud uploads, no internet connection required.

Latest models

Jumper uses the latest models and we update them regularly.

Key Features

Visual search

Search in footage using natural language keywords to retrieve relevant clips, no manual tagging required.

Instant Speech Search

Pinpoint specific spoken words or phrases across 111 supported languages, making it easy to locate dialogue in footage.

Face Detection & People Search

Instantly find every moment a person appears across your entire library.

Integrations

Jumper integrates within your editing software.

Unlimited & Offline

Jumper operates entirely offline, ensuring privacy and speed. No limit on analysis or usage.

Model updates

Access the latest ML models shortly after release.

Explore more

Quickstart

Get started with Jumper in minutes

The Interface

Master Jumper by exploring its interface and core features

Face analysis

Learn how Jumper detects, groups, and searches for faces

People tab

Name detected people and manage face collections

Summary Analysis

Give editors and agents a navigable overview of the footage

Agentic editing

Use Claude, Codex, or ChatGPT to orchestrate searches and deliverables
Last modified on July 29, 2026