AE GPT - AI Agent for After Effects
AI power. Creative flow.

AE GPT AI agent for After Effects

Ask for a scene, animation, effect, expression, or scriptable setup, and Agent Mode turns the request into real After Effects layers, keyframes, effects, and comps.

Get on aescripts - $29 Watch the demo
v1.2.5 - Agent Mode for AE
After Effects AI AI Agent Agent Mode Shape Layers ExtendScript Expressions Ollama
Composition-aware chatReads active comps, selected layers, masks, effects, properties, and expressions.
Scriptable executionTurns natural language into visible steps and runs After Effects automation where supported.
Flexible AI providersUse GPT, Claude, Gemini, through CLI or API keys, or private local Ollama models.
Product demo

See the panel, the plan, and the result.

A little preview of how AE GPT works inside After Effects, from getting natural language input to creating a new composition with layers, keyframes, effects, and expressions.

AE GPT panel running inside Adobe After Effects with Agent Mode steps
Main homepage GIFassets/aegpt-hero-demo.gif
Use your best full-screen After Effects + AE GPT capture.
Agent Mode in contextPrompt, execution plan, AE composition, and generated result in one visual.
AE GPT provider setup with Gemini OpenAI Claude and local Ollama options
Provider setup GIFassets/aegpt-provider-setup.gif
Choose your modelOpenAI, Claude, Gemini, CLI login, API keys, or local Ollama.
AE GPT chat mode with image references and voice input
Chat + voice GIFassets/aegpt-chat-voice.gif
Ask, attach, speakChat Mode, image/PDF references, and Whisper voice workflows.
Agent Mode - After Effects

It doesn't just answer. It executes.

Describe the result. AE GPT optimizes the prompt, plans the steps, writes ExtendScript, and creates After Effects elements live inside your workspace.

Agent Window

"create a new 4K comp and add a frog sitting in there with his eyes blinking"

OK Initializing 4K composition (3840x2160)
OK Building shape-layer frog scene
... Rigging eye-blink expression
04 Adding subtle breathing motion
05 Organizing layers and preview
AE GPT Agent Mode building a blinking frog scene with shape layers inside After Effects
Agent frog GIFassets/aegpt-agent-frog.gif
Replace this with your blinking frog capture.
Agent executionAE GPT writes layers, keyframes, effects, and expressions inside After Effects.
Bring your own intelligence GPT-5.5 CLAUDE OPUS / SONNET GEMINI 3.x OLLAMA - LOCAL
Chat Mode - After Effects first

A project-aware helper for expressions, effects, and motion ideas.

Ask questions, debug expressions, explain effects, generate scripts, and turn messy After Effects context into clear next steps. Premiere Pro chat support is a bonus for questions and guidance.

Chat - After Effects GPT / Claude / Gemini / Local
My project has an expression error. Find which layer contains the error and fix it.
AE GPT Found it on layer 13: "Logo Pulse Control" - Adjustment Layer. The Opacity expression is calling a slider named "Glow Amount," but the actual effect is named "Glow Strength."
What should the corrected expression be?
Expression fix Use effect("Glow Strength")("Slider") instead of effect("Glow Amount")("Slider"). I can also switch to Agent Mode and apply the fix directly in After Effects.
Message AE GPT...

Context-aware. Project-aware.

AE GPT lives inside your Extensions panel and can refresh After Effects context from the active comp, selected layers, masks, effects, properties, and expressions.

  • - Reads selected layers, masks, effects, comps, and expressions
  • - Helps write, debug, and explain After Effects expressions
  • - Generates ExtendScript for scriptable After Effects tasks
  • - Includes Premiere Pro Chat Mode as a bonus for questions and workflow guidance
Voice and references

Speak the idea.
Drop in the reference.

Use voice input for long animation prompts, then attach images or PDFs in supported API modes so AE GPT can follow a style frame, UI shot, palette, layout, script, or client note while working inside After Effects.

Voice input - Whisper

"add a slow camera dolly toward the frog over four seconds, ease-out, then hold"

Reference attachment
Why motion designers use AE GPT

Built for real After Effects work.

01

Native panel workflow

Runs inside the Adobe Extensions panel so prompts, project context, and creative decisions stay close to the timeline.

02

After Effects Agent Mode

Plans visible steps, optimizes the prompt, writes ExtendScript, and executes scriptable composition changes in After Effects.

03

Chat Mode for AE plus Premiere

Ask After Effects questions, troubleshoot expressions, explain effects, and use Premiere Pro chat as a bonus for questions and workflow guidance.

04

Reference-aware prompting

Add image or PDF references in supported API modes for moodboards, screenshots, layouts, client notes, and visual direction.

05

Voice and local transcription

Dictate animation notes, long prompts, and creative revisions with provider transcription or local Whisper setup options.

06

Cloud or local AI

Use OpenAI, Claude, Gemini, or local Ollama models depending on the job, budget, privacy need, and provider availability.

Models, login, and local AI

Choose the brain for each creative task.

AE GPT is provider-flexible: use API keys, supported account login through CLI tools, or local Ollama models for private local workflows.

OpenAI GPT-5.5, GPT-5.4, GPT-5.1, GPT-4.1

Strong for After Effects expressions, script generation, creative planning, troubleshooting, and detailed animation instructions.

API keyCodex CLIReferences in API mode
Claude Claude Opus and Sonnet models

Useful for long reasoning, breaking down motion design tasks, explaining code, and preparing reliable ExtendScript plans.

API keyClaude Code CLIAgent planning
Gemini + Local Gemini 3.x, Gemini CLI, Ollama

Use Gemini for fast multimodal workflows, or switch to local Ollama models when you want private prompts on your own machine.

Gemini APIGemini CLIOllama localhost
Real-world use cases

For motion graphics, compositing, expressions, and automation.

After Effects AI automation

Create compositions, organize layers, generate shape layers, build animation systems, add keyframes, write expressions, repair broken expression code, and explain project structure.

Expression and script help

Write wiggle, loop, bounce, slider-control, opacity, camera, and transform expressions, then generate ExtendScript for repetitive After Effects setup tasks.

Reference-driven creative direction

Use screenshots, images, brand boards, PDFs, and client notes as context for color palettes, layouts, visual systems, animation timing, and production planning.

AI workflow for motion designers

Replace scattered browser chats with a dedicated After Effects agent that understands composition context, model choice, and repeatable creative workflows.

AE GPT is designed for Adobe artists searching for an After Effects AI agent, After Effects AI assistant, motion graphics automation tool, expression helper, ExtendScript generator, local Ollama Adobe workflow, ChatGPT for After Effects, Claude for After Effects, Gemini for After Effects, and a creative assistant inside the animation workspace. Premiere Pro Chat Mode is included as a bonus for questions and guidance, while Agent Mode focuses on After Effects execution.

Details.

What is AE GPT? v
AE GPT is a context-aware AI agent panel for Adobe After Effects. It helps with project questions, expressions, scripts, references, voice prompts, and Agent Mode that takes over After Effects and directly executes what you ask of it.
Does AE GPT support After Effects and Premiere Pro? v
AE GPT focuses on Adobe After Effects. Agent Mode is for After Effects execution. Premiere Pro support is available as Chat Mode guidance for questions and project help, not direct video timeline editing.
What can Agent Mode do inside After Effects? v
Agent Mode can plan and run ExtendScript to create compositions, add shape layers, organize layers, apply effects, set keyframes, write expressions, and build scriptable animation setups inside After Effects.
What should I not expect it to do? v
AE GPT is not a direct video or audio editor. It does not cut footage, edit audio tracks, mix music, or replace a full NLE workflow. Its strongest execution workflow is scriptable After Effects work.
Which AI providers can I use? v
AE GPT supports OpenAI GPT models, Anthropic Claude models, Google Gemini models, and local Ollama models. Depending on the provider, you can use API keys, supported CLI account login, or local model detection.
Can AE GPT use images, PDFs, and voice input? v
Yes, with provider limits. Supported API modes can use image and PDF references. Voice input can use provider transcription or a local Whisper setup with tiny, medium, and large model options.
Can I run local AI with Ollama? v
Yes. Local mode can detect an Ollama service running on your computer and use installed local models. Image/PDF references and some cloud transcription features still depend on the selected provider mode.
Where do docs and support discussions go? v
Go to ae-gpt.com/docs/ for documentation, and ae-gpt.com/discussion/ for public comments, questions, issues, and user answers.