Plans your shots
Turns a one-line brief into a shot-level prompt — timeline, camera, and sound.

Brief it in chat. It plans, prices, and generates.
A conversational agent that runs image, video, and audio generation for you.
Turns a one-line brief into a shot-level prompt — timeline, camera, and sound.
Chooses from 20+ integrated models per task, or uses the one you name.
Every task shows an estimated credit cost before anything runs.
Inspects the images, video, and audio you reference — not just your words.
Like a creative director, it asks the right follow-up questions before spending anything.

It chains tasks in one thread: an anchor image first, then video shots from it, then sound.

Each task arrives as a card with prompt, model, and estimated credits — nothing generates until you confirm.


Type the goal — a product ad, a character, a storyboard. Attach images, video, or audio if you have them.

The agent replies with a task card: prompt, model, format, and the estimated credit cost.

Confirm the card to run it. Results return to the chat, ready to iterate, upscale, or download.
Renoise Agent is included with every plan during beta — credits are spent only on the generations you approve.
Renoise Agent is a conversational AI agent built into Renoise. You describe the goal in chat; it asks clarifying questions, writes shot-level prompts, picks a model, and generates images, video, and audio after you approve the cost.
Chat assistants write text. Renoise Agent executes: it inspects your uploads, drafts the prompts, chooses among 20+ integrated models, and actually runs the generation — with the credit cost shown before anything happens.
Yes — it plans multi-step work on its own: reference image first, then video shots built on it, then sound, all chained in one conversation instead of separate manual generations.
No. A one-line brief is enough — the agent expands it into a detailed prompt with camera, timeline, and sound direction. You can still edit the prompt on the task card before it runs.
No — conversation does not consume credits. You pay only for generations you approve, at the estimated credit price shown on each task card. Approval can be set to Always, Ask on Credits, or Auto.
The models integrated in Renoise: Seedance 2.0 and Kling 3.0 Omni for video, Seedream 5.0, Nano Banana Pro, and GPT Image 2 for images, Seed Audio 1.0 for sound — or exactly the model you name.
Yes. It generates an anchor image first, then reuses it as a shared reference across variants and follow-up shots — a workflow that noticeably improves character consistency.
It can restyle footage with video-editing models and run upscale, repaint, and erase on results, but it does not cut or stitch clips — multi-shot sequences are planned into a single generated video instead.
Renoise Agent is in beta and included with any active Renoise subscription at no additional cost. It currently runs on desktop web.
No. Renoise Agent is the creative agent inside renoise.ai. The agent-ready APIs are separate — they let coding agents like Claude Code or Codex call Renoise from outside via API and MCP.
Included with every Renoise subscription during beta.