/mango_tango/v1.0/manifesto.html●●●signed · 04.26.2026 · park_city_ut
manifesto // v1.0
[[signed]][read once · share once]

a film tool.
not a slot machine.

this is what we believe. this is what we reject. this is who we built the product for. read it once. if it sounds like you, the rest of the site will too.

// thesis

the last two years gave us magic and waste in equal measure. magic: you can describe a shot, and a model will draw it, light it, animate it, score it. waste: most of the products built around that magic treat creators like they're playing a slot machine — pull the lever, see what comes out, scroll the gallery, pull again.

that's not how films get made. films are decisions stacked into sequence. shot, cut, shot, cut, shot. the smallest unit that matters is the cut, because a cut implies a story. a tool worthy of this technology should put the cut at the center, not the generation. that's the entire premise of mango tango.

// section_01 · what_we_believe

nine positions.
no flexibility.

01

the cut is the unit.

ai video has been treated as a slot machine — type a prompt, get a clip, scroll the gallery, repeat. that's not cinema. that's content. the smallest meaningful unit of a film is the cut, not the shot. a cut implies sequence, intent, structure, story. mango tango is built around the cut. generation is a button on the timeline, not a destination you bounce to.

02

generate from intent, not from boredom.

every clip in your project should exist because you needed it for the cut, not because the model produced something pretty enough to keep. the gallery model rewards aimless scrolling. we reward decisions. you start with the story, you generate to fill the gaps, you trim until it works.

03

consistency is the moat.

your protagonist should hold across thirty shots. your product should look like itself in every hero frame. character and product references are first-class objects in mango tango — named, draggable, attached to a clip range, propagated automatically. every other ai video tool buries this in a settings dialog. for us, it's the spine of the product.

04

the timeline is the home.

the canvas where clips become a film is the center of the application. not a sidebar. not a separate workflow. when you log in, you land in a project. when you generate, you're already on the timeline. preview the assembled cut at any time without context-switching. that's the only way the unit-of-work matches the way films are actually made.

05

pay for what you ship.

we do not run unlimited tiers. we never will. unlimited is theater — every company offering it is either burning vc money or hiding caps in fine print. we sell credits. one credit equals one unit of work. the price is on the page. you can never owe us money you didn't agree to spend, and we can never be tempted to bury usage limits in terms of service. the math is the menu.

06

private by default.

your work is your work. every project starts private. it stays private until you flip a toggle. your generations are not training data — not for us, not for the upstream models we route through, not for anyone. when we say private, we mean an end-to-end commitment to keeping your unfinished projects unseen.

07

tools should look like tools.

we reject the purple-gradient saas aesthetic that has become the default uniform of every ai startup. tools should feel like tools — heavy, opinionated, structurally honest. monospace type. visible borders. no rounded corners. no fake friendliness. you are doing real creative work. the chrome should respect that.

08

constraints are a feature.

one video track. three audio tracks. as many clips as the story needs. constraints force focus on what matters — the cut, the pacing, the read — without arbitrary clip caps getting in the way of a longer edit when you have one.

09

make it move.

the entire product reduces to two words. you came here with an idea — a script, a treatment, a one-line concept, a vision in your head. by the time you leave, that idea is moving. it's a thirty-second cut you can post, pitch, send, or print onto a deck. that transformation is what we sell. everything else is plumbing.

// section_02 · what_we_reject

and equally,
what we don't do.

a brand is what you say no to. these are the choices we made by refusing the alternatives. no asterisks, no exceptions, no quiet rollbacks when growth gets hard.

// section_03 · who_this_is_for

built for the
people who haven't shipped yet.

we don't care if you've directed a feature. we care that you have something you want to make. mango tango is built specifically for these people:

// audience_01
the screenwriter
with a finished spec script, a notebook of scene notes, and no crew. who needs to see act one as a sizzle before sending the pdf to a manager.
// audience_02
the music video director
burning self-funded weekends. who's been quoted $40k for a treatment she could now ship herself in a long saturday.
// audience_03
the indie filmmaker
pitching out of a coffee shop. who needs proof of concept on screen before anyone will write a check.
// audience_04
the brand marketer
running eight simultaneous campaigns. who's tired of three-week turnaround on a six-second clip.
// audience_05
the small-agency creative
who knows premiere cold but is bored. who wants to spend creative energy on the cut, not on render queues.
// audience_06
the wannabe
the person who has been calling themselves a filmmaker in private for years and needs to see one finished thing before they say it out loud.
// section_04 · signed_and_dated

this is the
contract.

if mango tango ever drifts from the nine positions above — if a future version starts treating users like prompt typists instead of filmmakers, if the unit becomes the clip again instead of the cut, if “unlimited” ever appears on a pricing page — this manifesto is the receipt. screenshot it. mail it back. hold us to it.

we will not change these positions because growth is hard or a board meeting demands quarter-on-quarter expansion. the day the product stops respecting the cut is the day we're no longer building mango tango. it'll be something else with the same domain.

// signed
xavier lampkin
// role: founder
// date: 04.26.2026
// location: park city, ut
// domain: mangotangoai.com
the cut is the product. ship something_
start_creatingsee_what_we_built