Live now ChatGPT app — coming soon
MCP server

Manage your films and videos from your AI assistant

Connect Vidreno to ChatGPT, Claude, Cursor or any MCP client and manage your films, projects and hosted videos by chatting: creating a film, its release details and prices, publishing with the editor's checklist, crew credits, posting a project and answering its applicants, finding open roles and applying, your earnings and payout history (read-only), embed codes with the right player options, analytics, uploads, even deletions (with your explicit yes). The MCP server is live for Claude, Cursor and any MCP client; the ChatGPT app is coming soon.

  • Included with every creator account
  • Films, projects, earnings and hosted videos
  • You approve what it may do
  • Disconnect at any time
The basics

What is MCP?

Three things worth knowing before you connect anything.

An open standard

The Model Context Protocol (MCP) is how AI assistants use outside tools. A server describes what it can do; the assistant calls those tools for you.

Nothing to install

Vidreno's server is remote. Your assistant connects to it over HTTPS at https://vidreno.com/mcp, signs you in through Vidreno and keeps a token that covers only what you approved.

Real answers, not guesses

Ask why your film won’t publish and the assistant reads its real checklist; ask for the embed code for the launch teaser and it returns the real snippet, with the player options you asked for.

What you can ask

Plain requests, real actions

Each example is something you can type as-is. Behind it, the assistant calls one of the server’s tools on your account.

Films · Read

Find a film

“What's the status of my short film, and how many plays has it had?”

List your films, pull up one film's status, how it's released (free to watch, or buy or rent) and prices, play count, credits, main film and trailer state, publish checklist and public page link. Films are never embeddable, so there is no embed code to ask for.

list_filmsget_film
Films · Write

Create a film

“Create a new film called Night Shift”

A new draft from just a title, as in the Studio: the assistant gets the film's id, its editor link and its publish checklist straight back, then fills in the details, points you to the upload link for the main film and to a browser handoff for the poster, and publishes when the checklist is complete.

create_film
Films · Write

Release details and prices

“Make Night Shift free to watch and update the synopsis”

Title, synopsis, director and cast lines, year, runtime, genres from the fixed list, country of production, what it was shot on, student film, filming locations, awards (checked by the team before they show), IMDb link, extra links (press kit, festival page), free to watch or buy or rent, and buy and rent prices (from $1.00). Main-film, trailer and artwork uploads use browser handoffs; nothing an assistant does removes a file. The same validation as the film editor; only the fields you mention change.

update_film
Films · Write

Publish, with the editor's checklist

“Publish Night Shift”

Publishing goes through the film editor's checklist (title, poster, a ready main film, and a price only when it's buy or rent; a free film needs none); if something is missing the assistant gets the same message you would and tells you what to fix, with the editor link for the poster, which is added in the editor. Unpublishing takes the film back to draft.

publish_filmunpublish_film
Films · Write

Crew credits

“Add Sam Lee as composer and invite them by email”

Add, edit, remove and reorder credits, and turn older director and cast lines into credit rows. Link a credit to a filmmaker by username (they accept it on Vidreno) or invite the person by email, or ask for an invite link to pass on yourself, with the same limits as the app. Editing keeps an accepted credit accepted.

add_film_creditupdate_film_creditremove_film_creditreorder_film_creditsinvite_film_creditconvert_legacy_credits
Profile · Write

Your filmmaker profile

“Update my bio and add Letterboxd to my profile, then accept the credit Sam gave me”

Read and edit the profile your public page is built from: username, display name, bio, location, roles, website and social links, and whether it is public, with the Profile page's rules and messages. See the credits other filmmakers gave you and accept, decline or unlink them. Avatar uploads use a browser handoff.

get_filmmaker_profileupdate_filmmaker_profilelist_my_creditsrespond_to_credit
Films · Write

Upload a main film or trailer

“Upload the new cut as the main film”

Same one-time browser link as a hosted video, aimed at the film's main film or trailer slot. The file goes to film storage (outside your plan storage), replaces the current one when it finishes and the assistant checks the film afterwards.

start_film_uploadget_upload_status
Earnings · Read

Earnings and payouts

“How much have I earned this month, and when was my last payout?”

What your Payouts page shows, read-only: your balance (available, maturing, paid out), lifetime and this month's film-sales and ad earnings, a per-film breakdown, every order with Vidreno's cut and the processor fee (never who bought), your payout history and your payout method in masked form. Cashing out and changing where you are paid stay on the Payouts page.

get_earningslist_payoutslist_film_sales
Projects · Write

Post a project and answer applicants

“Draft a crew call for the short we're shooting in Manchester in November, with a paid gaffer and two runners, and publish it”

Create a project (public, or private and invite only), fill in its details and dates, add roles with the pay stated on each, see the Studio's publish checklist and publish it; then read the applications with each applicant's public profile card and note, shortlist, accept or decline with a message, reply in the thread, mark a role filled and close the project when you are done. Applicants are told by e-mail exactly as if you had used the Studio. Invite collaborators, upload a project cover through a browser handoff, and read or contribute to permitted discussions.

list_projectsget_projectcreate_projectupdate_projectadd_project_roleupdate_project_roleremove_project_rolepublish_projectclose_projectlist_project_applicationsrespond_to_applicationsend_application_message
Projects · Write

Find open roles and apply

“Any paid camera or lighting roles in the UK I could apply to? Apply to the best one with a short note”

Search every open role across live public projects by text, department or role, country, remote-friendly, pay type and project type, with the pay stated on each role and the project owner's public profile card, then apply as you (with a note and links), follow your applications, message the owner and withdraw. The owner sees the same application and notifications as one sent from the Jobs page.

search_open_rolesapply_to_rolelist_my_applicationssend_application_messagewithdraw_application
Hosting · Read

Find a hosted video

“Which of my videos mention the spring launch?”

Search your library by title, filter by public or private, and pull up one video's details, links and view count.

list_videosget_video
Hosting · Read

Embed code with player options

“Give me a responsive, autoplay-muted embed code for launch-teaser.mp4”

Autoplay, muted, loop, controls, corner radius, fullscreen, responsive or fixed size and alignment: the same options as the embed modal, ready to paste.

get_embed_code
Hosting · Write

Housekeeping

“Rename it to “Launch teaser (final)” and make it private”

Title, description, visibility, player colour and category. Only the fields you mention change.

update_video
Hosting · Read

Analytics

“How are my videos doing this month?”

Totals, your most-viewed videos, the newest uploads and a daily views series for the window you ask about.

get_analytics
Hosting · Write

Upload a video

“Upload a new video called Behind the scenes”

Ask for an upload and the assistant hands you a one-time link: you pick the file in your browser and it goes straight to Vidreno. In ChatGPT you can also attach a video to the chat (up to 512 MB) and Vidreno fetches it from the attachment. Either way the assistant reports back with the finished video and its embed code.

start_video_uploadupload_video_from_chatget_upload_status
Hosting · Write

Hide it, never delete it

“Take the old teaser down”

An assistant cannot delete a video. It makes it private, which you can undo; deleting stays with you, on the video's own page.

update_video
Films · Write

Prepare film files and subtitles

“Check my film processing and subtitle progress”

Choose structured languages, inspect processing stages, upload artwork and masters through your browser, manage bonus materials and subtitle tracks, and publish ready replacement versions. Automatic subtitles appear only where enabled.

get_film_processingset_film_languagesstart_film_file_uploadstart_film_master_uploadlist_film_subtitleslist_film_bonus
Account · Write

Manage creator preferences

“Show my branding and account preferences”

Read or update your player branding, creator mode and marketing preference. View bandwidth and connected apps. Sign-in details, billing, payouts and app disconnection stay in the website.

get_brandingupdate_brandingget_account_settingsupdate_account_settingslist_connected_apps
Connecting

How to connect

Two minutes, no API keys. The AI app opens a Vidreno sign-in, you approve what it may do, and the connection is made.

ChatGPT

Coming soon
  1. 1

    When the ChatGPT app launches, find Vidreno in ChatGPT's apps and choose Connect.

  2. 2

    Sign in to your Vidreno account, or create a free one on the spot.

  3. 3

    Approve the consent screen. It lists exactly what the assistant may see and change, and you can disconnect it from the AI app at any time.

  4. 4

    Ask away. ChatGPT uses your Vidreno account only for what you ask in the chat.

Until then, the server already works with any MCP client that supports remote servers; follow the steps in the other card.

Claude, Cursor and any MCP client

Connect now
  1. 1

    Add https://vidreno.com/mcp as a remote MCP server in your client.

  2. 2

    Sign in to your Vidreno account, or create a free one on the spot. Your client registers itself and runs the sign-in flow in your browser; there are no API keys to paste.

  3. 3

    Approve the consent screen. It lists exactly what the assistant may see and change, and you can disconnect it from the AI app at any time.

Any client that speaks the Model Context Protocol over HTTP with OAuth works the same way; Claude and Cursor are two examples.

Server URL for any MCP client

https://vidreno.com/mcp
Step-by-step guide
Safe by design

What it can and cannot touch

The server is built so an assistant can help with your videos without becoming a key to your whole account.

Scoped to you

Every call runs as the account you signed in with and can only reach that account’s films, projects, earnings and hosted videos. Other people appear only as their public profile cards: the applicants to your projects, and the owners of the open roles you search.

Your films, projects, earnings and hosted videos, nothing else

Projects: it can post and manage your own crew calls, answer applicants and apply to open roles as you, with the Studio’s own rules and e-mails; it can also manage project invitations, upload covers through browser handoffs and participate in permitted discussions. Earnings, film sales and payouts are read-only and behind their own permission: it can see them (never who bought, never your full payout details) but never cash out or change where you are paid. No billing, advertiser tools or admin access. Those stay in the dashboard, and films are never embeddable.

Read or write: you decide

11 permissions: read or write for films, your filmmaker profile, hosted videos and safe account preferences, plus projects, plus read-only access to earnings and payouts. The consent screen names each permission. A read-only connection cannot change anything.

videos:read
See your hosted videos, their settings, embed codes and analytics
videos:write
Rename, describe, change the visibility of and upload videos, and set account player branding (it cannot delete them)
films:read
See your films: status, how each is released and priced, play counts, credits and the public links
films:write
Create and edit your films, change how they are released, publish and unpublish them, manage credits and invites, and upload film files
profile:read
See your filmmaker profile (username, display name, bio, location, roles, website and social links, whether it is public) and the credits other filmmakers gave you
profile:write
Edit your filmmaker profile, including its username and whether it is public, and accept, decline or unlink credits other filmmakers gave you
earn:read
See your earnings, balance, film sales and payout history, and your payout method in masked form (read-only: it can never request a payout or change where you are paid)
projects:read
See your projects (roles, publish checklist, status, links), the applications to them with each applicant's public profile card, notes and messages, your own applications, search open roles, read invitations and permitted project discussions
projects:write
Create, edit, publish and close your projects and their roles, shortlist, accept or decline applicants and message them, and apply to open roles, message and withdraw as you; send, revoke and respond to project invitations, post in permitted discussions, manage your own project posts and notifications
account:read
See creator preferences, marketing email preference, bandwidth usage and connected apps (without access tokens)
account:write
Change creator mode and your marketing email preference; cannot change sign-in email, password, billing, payouts or connected apps

A consent screen first

Connecting always shows, on Vidreno, what the app may see and change before any access is granted.

Disconnect at any time

Remove the app in your AI assistant and its access ends. Tokens expire on their own anyway: access after 1 hour, refresh after 30 days.

Uploads take the short path

Browser uploads go straight from your device to Vidreno through a one-time link; in ChatGPT you can also attach a video to the chat and Vidreno fetches it from the attachment. Either way the assistant only learns the result, and video bytes never travel over the MCP connection itself.

Questions

MCP server FAQ

Do I need a paid plan to connect an AI assistant?
No. Connecting is included with every creator account, the Free plan included. Your plan's storage and bandwidth allowances apply to your videos as usual, however they were uploaded.
When can I connect ChatGPT?
Soon. The ChatGPT app is not available yet; the server itself is already live for Claude, Cursor and any MCP client, so you can connect from one of those today.
Which AI assistants work with it?
Any client that supports remote MCP servers with a sign-in: ChatGPT (as the Vidreno app), Claude, Cursor and any MCP client. The steps are the same everywhere: add the server, sign in to Vidreno, approve what it may do.
Can the assistant delete my videos?
No. Nothing an assistant can do deletes a video, a film, a season, an episode, a subtitle track or any file. You delete those yourself on Vidreno. What the assistant can do instead can be undone: make a hosted video private, unpublish a film, season or episode, or switch a generated subtitle track off.
Can it see my sales, earnings or payouts?
Only if you grant the separate, read-only earnings permission when you connect, and then only what your Payouts page shows: your balance, film-sales and ad earnings, each order's price, Vidreno's cut and the processor fee (never who bought), your payout history and your payout method in masked form. It can never request a payout or change where you are paid; that stays in the app. Beyond that the server reaches your films (creating a draft, details, release mode and prices, publishing, crew credits, uploads), your projects (crew calls, applicants, open roles) and your hosted videos, and nothing else: billing and payment details, advertiser tools and admin features are not exposed to it at all. Films are never embeddable, so there is no film embed code either.
Can it change my public filmmaker profile?
Only with the separate profile permission. Reading covers what your Profile page shows (username, display name, bio, location, roles, website and social links, whether the profile is public) and the credits other filmmakers gave you; writing lets the assistant edit those fields, with the same rules and messages as the Profile page, and accept, decline or unlink a credit. Your avatar, film posters and deleting a film stay in the app.
Can it post my projects and answer applicants?
Only with the separate projects permissions. Reading covers your own projects (roles, status, publish checklist, applications with each applicant's public profile card and note, the message threads), your own applications, and the public open-roles search. Writing lets the assistant create, edit, publish and close a project, add and change its roles and pay, shortlist, accept or decline applicants with a message and reply to them, and apply to open roles as you, with the same rules, caps and e-mails as the Studio; closing a project needs your explicit confirmation. Inviting people, cover images and discussions stay in the Studio, and no tool returns anyone's e-mail address.
Does my video file go through the AI?
The assistant hands you a one-time link. You pick the file on your device and it goes straight to Vidreno; the assistant never handles the file and is only told when it is done. In ChatGPT you can attach a video to the conversation (up to 512 MB) and ask to store it. That file is already held by ChatGPT; Vidreno fetches it from the attachment and the assistant only receives the result. In both cases the video never travels over the MCP connection itself, and the assistant only gets the finished video's details and embed code.
How do I disconnect?
Remove the Vidreno app in your AI assistant's app or connector settings and its access ends there. Any tokens it held expire on their own: access tokens after 1 hour, refresh tokens after 30 days.
Is there documentation for developers?
Yes. The server speaks standard MCP over HTTPS and signs users in with OAuth; clients register themselves, so there are no API keys. The public discovery document lists the endpoints and scopes. Read the discovery document

Building an MCP client or an agent? The endpoints, scopes and sign-in flow are documented at https://vidreno.com/auth.md.

Talk to your video library

The MCP server is live for Claude, Cursor and any MCP client; the ChatGPT app is coming soon.