Monte
Year
2026
Film
Announcement film
Launched
[date]
Monte
Year
2026
Film
Announcement film
Launched
[date]
The idea
Most AI agents run on the same generic models, so teams keep rewriting prompts, adding tools and rebuilding the harness to get better results. Monte’s founders, Alex Fleury and Jamin Liu, argue that open-weight models have changed that: a model trained on a company’s own knowledge and data can do its specific work better, at a fraction of the cost. The film is built around that one shift.
The craft
Shot in a wood-panelled library, with the two founders speaking straight to camera. The platform is drawn as a calm, step-by-step diagram on deep navy: the agent you run today, the harness, traces and knowledge that become environments and datasets, a training run climbing towards its target, and the loop of production traces and user feedback that keeps the model improving.
Where it ran
Delivered as a 1:16 launch film in 16:9, with a 3:4 cut for feeds, subtitled throughout and ending on trymonte.ai and a call to book a demo.