StarZero

A whole studio.
At your direction.

Find the story in your library. Edit in chat. Generate what comes next. Then run the same production on every new video your team publishes.

Talk to sales Security and data
Enterprise

Your production, as steps.
Run on every new video.

Workflows write an edit once, with variables, branches and renders, and run it on every episode. Here is a real run: one 20VC episode in, 8 clips out, every branch rendered in about 22 minutes. From 2,000 edits a month to more than 50,000.

See Workflows

NVIDIA’s 70% guidance shock and the hyperscaler cash-flow squeeze Straight from the run above, with no edits after render.

From one 20VC episode

The same template on 8 episodes

69 clips. On each episode the model decided how many moments were worth cutting, from 6 to 12, and the clone step made that many branches.

  1. 29 Sep 20VC

    NVIDIA Crushes Quarter | OpenAI Cuts Off Cursor

    8 clips

  2. 21 Aug 20VC

    ElevenLabs: Building an AI Sales Machine

    12 clips

  3. 5 Aug 20VC

    Coinbase Cuts AI Spend by 50% | Kalshi’s $40B Valuation

    6 clips

  4. 21 Jul

    How I Plan, Build, and Run Loops with Claude Code · Thariq Shihipar

    6 clips

  5. 16 Jul

    Grant Sanderson (3Blue1Brown): AI disproved a famous math conjecture

    9 clips

  6. 30 Jun

    Training Sand to Think: AGI & the Future of Physics

    9 clips

  7. 26 Jun

    NVIDIA GTC Keynote 2026

    9 clips

  8. 25 Jun

    From Revolut employee #3 to a £5B energy company · Alan Chang

    10 clips

Security and data.
We put it in writing.

Here’s what we commit to today. The full terms are in our Privacy Policy, DPA and sub-processor list.

Audit logs

Every Enterprise account includes audit logs.

Deletion and logs

When you delete media, we remove it from our systems within 90 days, backups included. We keep operational logs for 180 days.

Privacy Policy, section 8
Where it runs

StarZero processes your data primarily in the United States.

DPA, section 10
AI providers

The AI providers we use to analyze your media work under terms that bar them from training their models on it.

Sub-processor list
Faces

We discard the face embeddings used to tell speakers apart when each job finishes.

DPA, section 13
Breach notice and audits

We’ll notify you of a breach without undue delay, and you have the right to audit.

DPA, sections 8 and 11

We connect your sources.
Each new video starts a run.

Tell us where your videos come from. We connect them, and every new one is ingested and starts its Workflow run on its own. Results go back to your systems through a callback.

  • YouTube channels

    The channel notifies StarZero the moment a video goes live.

  • Social accounts

    New posts come in as they are published.

  • Storage buckets

    A new file in the bucket is a new video to run.

  • Web pages

    A page that lists episodes feeds in each new one.

  1. Ingested

    Transcribed, described and searchable in your library.

  2. Workflow run

    Your template cuts, reframes, captions and renders.

  3. Back to you

    Videos, titles and variables posted to your callback URL.

Every plan

The MCP server, the CLI and the Claude plugin, for search, clips, renders and the agent from your own tools. The REST API is available on request.

Developers

Enterprise

Workflows, connected sources, audit logs and a sales contact. Start runs from the API, or from MCP with a Workflow’s ID.

Pricing

Bring the operation.
We’ll start there.

Tell us where your videos come from, how many arrive each week and what each one should become.

Talk to sales