Producers say yes to what they can see. Instead of explaining your tone in a logline, hand them ninety seconds of it — cast, world, and camera language already on screen. Dreamz builds the trailer from your description in chat, before a single dollar of production budget exists.
Describe the mood — Coen-dry, slow-burn dread, neon melodrama — and the trailer's lighting, pacing, and score commit to it, so financiers grade the vision, not your prose.
You approve your leads from character portraits before filming, and they stay consistent through every scene of the trailer.
The trailer is cut from actual dramatized scenes of your story with native voices and lip-sync, so it plays like footage from a film that already shot.
The finished trailer is commercially yours to screen at markets, attach to applications, and send to every producer on your list.
Dreamz is also a Claude connector (MCP) — describe your film to Claude and the finished video comes back in the conversation. If Claude is already your workspace, your film studio now lives there too.
Tell Dreamz about the proof of concept trailer you want — story, mood, who's in it. The director agent writes the script and casts characters with you, in chat.
Scenes generate live on your dashboard — visuals, voices, music and editing handled for you, with the exact price quoted before anything is charged.
Watch the finished cut, request changes in plain words, then download it — everything you make is commercially yours.
You have thirty pages of a script that is genuinely good and a tone that is genuinely specific, and you are trying to transmit both through a sentence. It does not work, and the reason is not that your sentence is weak. Prose and film encode tone differently. "A darkly funny thriller about grief" is a true description of at least four hundred projects, half of which are unwatchable, and the person reading it has met all four hundred. They are not being lazy when they ask what it looks like. They are asking the only question their job permits: can this director hold a tone for ninety minutes. Historically the answer required money — you shot a proof of concept short on favors, burned a weekend of your DP's goodwill, and hoped the one location you could beg looked like the world in your head. Plenty of good films exist because that gamble paid. Plenty of good films do not exist because it did not. The proof of concept has always been the most effective document in independent film and the most expensive to produce, and that gap is where projects quietly die.
You do this in a chat, in plain language, the way you would talk to a producer who already read the script. Dreamz asks what it needs — runtime, aspect ratio if you did not say, whether it's one continuous scene or a cut sequence. Then it writes the screenplay and shows it to you before a single frame is rendered, which is the point at which you catch that it softened your ending or misread the joke. You revise in plain words. It quotes a price in chat, that quote is exactly what you are charged, and nothing bills until you approve it. A workable brief reads like this: "Ninety seconds, 2.39:1. Proof of concept for a feature called SALT LINE. A Portuguese-speaking crab fisherman in New Bedford, Massachusetts, February, working alone because his brother is in prison. Three scenes: hauling pots in grey pre-dawn light, a wordless dinner with his brother's wife, and a final scene where a Coast Guard boat pulls alongside and he keeps working without looking up. Dry, cold, almost no score. The tone is Kelly Reichardt, not Deadliest Catch. No trailer voice, no title cards until the last four seconds." Note what that brief does: it names the negative space. What it must not become is as load-bearing as what it is.
Most first-time proof of concepts illustrate the plot, which is the wrong job. Nobody funds a plot; plots are cheap and your logline already delivered it. What gets funded is evidence that a specific sensibility exists and can be sustained. So build the piece around the hardest thing in your film — the thing a reader would flag as a risk. If your feature depends on a tonal swerve from comedy to violence in one unbroken take, that swerve is your proof of concept, even if it happens on page 68. If it depends on a child actor carrying silence, film the silence. The instinct to open with your best action beat is usually wrong, because action reads as generic and doubt does not live there. Length discipline matters more than people expect. Ninety seconds of one committed scene beats three minutes of highlights, because highlights secretly signal that you could not sustain anything. And name your comparison honestly in the brief. Saying "Reichardt, not reality TV" gives the direction something to aim at and gives you a standard to reject the result against.
The finished piece is commercially yours, which means the distribution question is a strategy question rather than a rights question. Resist posting it publicly first. A proof of concept spends its value the moment it becomes a thing people have seen, and its highest use is as a private argument sent to a named person with a named ask. Attach it to a grant application where the guidelines permit visual support. Put it in the first two minutes of a market meeting rather than the last. Send it to the one producer whose slate actually contains your film, with a sentence saying what it is — a proof of tone, not a cut of the movie — because misrepresenting it as footage is the fastest way to lose the room. Be straightforward that it is AI-generated when you share it; distribution platforms and funding bodies increasingly have disclosure requirements and those policies change, so check the current rules of whoever you're submitting to rather than assuming. Honesty here costs you nothing. The piece was never pretending to be the film. It is pretending to be the future, which is exactly what a pitch is.
Everything on screen is generated for your story — your characters, your scenes, your dialogue. There is no borrowed footage to clear and no other film's fingerprints on your concept.
Dreamz quotes the exact price in chat before anything is charged, and the approved quote is the exact charge. Short videos start at a few dollars, so you can test a teaser-length cut cheaply first.
Yes. Paste or describe the scene in chat and Dreamz stages it faithfully — the dialogue, the beat, the reveal — rather than inventing around it.
Yes. Characters are cast once, approved by you from portraits, and held consistent across the whole trailer, which is exactly what proof-of-concept audiences look for.
Yes. Dreamz works as an official Claude connector (MCP), so you can develop the deck and commission the trailer in the same workflow.
Yes, plainly and up front. Present it as a proof of tone rather than footage. Festivals, funds, and platforms are actively developing AI-disclosure requirements and those rules differ by body and change often, so check the current policy of whoever you are submitting to before you attach anything.
The one that carries the most doubt, not the one you like best. If readers keep asking whether the tonal shift works or whether the premise sustains, film that. A proof of concept exists to close a specific objection, so pick the objection first and let it choose the scene.
Yes. You continue in the same chat, so the cast you approved and the world you established carry forward into a longer piece. Each new film is quoted separately in chat before anything is charged, so extending the concept never commits you to a price you have not seen.
Yes. Dreamz ships an official Claude connector (MCP): connect it once in Claude, then describe the video you want in any conversation — pricing quotes, drafts and the finished film are delivered right in the chat. The website and the Claude connector share the same account and pricing.
Describe your story in chat and get a trailer that argues for the budget.