Problem

Linum is tackling the cost and complexity of making narrative and animated video. The company argues that financing is a limiting constraint on filmmaking: producing a movie is expensive, while raising money to produce one is difficult. Animation remains especially inaccessible; even an acclaimed independent film such as Flow was cited as costing roughly $3–4 million. The killer use case is enabling filmmakers—particularly independent animators and storytellers—to move from an idea to a first visual sequence without the conventional production budget.

The company also targets the usability gap in traditional animation software. Tools such as Blender are powerful but difficult to operate because they are “functionally rich” yet semantically poor: users can do almost anything, but expressing a specific creative intent requires substantial technical work. Linum’s vision is to make high-quality animation more accessible through natural-language and other intuitive creative controls, potentially reducing production costs by an order of magnitude.

Product / Service

Linum is a small research lab that trains its own generative media models from scratch. Its January 2026 Linum v2 release consists of two open-weight, 2-billion-parameter text-to-video models that generate two-to-five-second clips at up to 720p. The models can be downloaded and run on a local machine, with the release distributed under the Apache 2.0 license alongside source code, making the delivery model closer to an open model platform than a conventional hosted video SaaS product.

The models convert text prompts into video and are intended to provide more semantically meaningful creative control than traditional animation workflows. Linum says that building a video-native system gives it control over the model and dataset rather than trying to retrofit an image model for video. The current release is still an early foundation: generating a five-second 720p clip takes about 15 minutes on an H100, and the company identifies physics, aesthetics, deformations, speed, audio, and scaling as areas for continued development.

Market

Linum competes in generative AI software, specifically the text-to-video and AI animation market. Its competitive field includes closed-source systems such as OpenAI’s Sora and Kuaishou’s Kling, as well as broader AI video products such as Runway. Linum’s differentiation is its open-weight, downloadable approach and its longer-term focus on making narrative animation and filmmaking more accessible, rather than only offering a consumer-facing video-generation interface.

It is not best described as pre-revenue based on available third-party estimates: GetLatka reports approximately $5 million in estimated 2024 revenue or ARR, a $15 million valuation, and a two-person team. Those figures are not corroborated by Linum’s own 2026 product announcement, which describes the company as a two-brother research lab and emphasizes model releases rather than customer counts; therefore, revenue should be treated as directional. The clearest verified current traction is the public release of Linum v2’s 360p and 720p models, weights, and source code, while public evidence does not establish a customer total.