Problem

Martini addresses the mismatch between how filmmakers work and how generative-video tools typically work. Existing systems often reduce direction to prompt-and-seed experimentation, while professional filmmakers think in terms of intentional shots, camera placement, lenses, movement, and collaboration. The result is “prompt roulette”: creators may spend substantial time iterating without reliably getting the precise shot they want.

The pain is also operational. Production teams are forced to stitch together multiple generation, editing, and collaboration tools, creating fragmented workflows and coordination overhead. Martini’s primary use case is therefore collaborative cinematic production—especially for studios, agencies, and filmmakers creating commercials, gallery work, social content, or other short-form films that require repeatable shot direction rather than one-off AI clips.

Product / Service

Martini is a cloud-based AI filmmaking platform that sits on top of multiple third-party image and video models rather than training its own models. It gives users a virtual film set with camera-position, lens, and movement controls; lets them enter generated or uploaded images and video to reframe and reshoot; and supports composition in virtual 2D or 3D environments. The benefit is a more cinematography-native way to direct AI output and explore shots.

It also provides a production layer for teams: real-time sharing of prompts, edits, and feedback; a built-in timeline for rough assemblies; and XML export into professional editing software. Martini is delivered as a web subscription and usage platform. Its current pricing uses “Olives,” with one Olive equal to $0.10 and generation charged by the second across available models; the site lists a free Indie plan, a $30-per-month Pro plan for solo professionals, a $150-per-month Studio plan with team controls, and custom Enterprise plans.

Market

Martini competes in the emerging market for professional AI-video production and filmmaking workflow software: a layer above raw generative models that aims to bring cinematography, editing, and collaboration into one environment. Its adjacent and direct alternatives include Runway and LTX Studio, as well as newer production-oriented platforms such as Electric Sheep, Arcana Labs, and MovieFlo.AI. Martini differentiates around camera control and collaborative production rather than simply offering another text-to-video generator.

The company appears to be early-stage but commercially launched. Martini says more than 200 films have been made with the platform, including TV commercials, gallery artwork, and viral online content; it is listed by Y Combinator as a Winter 2026 company founded in 2025 by Long Hoang and Koh Terai, with a two-person team in San Francisco. Live free and paid plans establish a monetization path, but the public materials reviewed do not disclose revenue, paying-customer counts, or funding, so the available evidence supports early traction rather than proof of scaled commercial adoption.