“just use a foundation model” they said. so we benchmarked 10 of them against Manako on a real forecourt damage clip. same full 1280x720 frames, same class prompts (person, car, fuel dispenser, fuel nozzle), no crops, no zoom. the hard class is the nozzle/hose. small, thin, occluded, and the thing that actually gets damaged. - Grounding DINO labelled a bin as a fuel nozzle. - Florence-2 labelled the canopy pole as one. - OWL-ViT, YOLOE and YOLO-World got the dispenser at best, never the nozzle. - RF-DETR doesn’t have the class. - Gemma 4 E2B described the scene, boxed the pump, and returned no incident time. the only open model that localised the nozzle was full SAM3. 3.45 GB, 0.2-1 FPS on a Jetson Orin Nano. Manako (3 specialists - 29.5 MB), 3-6 FPS on the same Jetson, tracked the nozzle at 5 FPS and fired a live equipment_damage event on the hose drop at 10.3s. 117x smaller than the only model that could keep up. open-vocab models are great at the classes the internet photographs a lot. operations live in the long tail: nozzles, hoses, cabinets, the small parts that break. that’s where specialised detectors win, and it’s why we built Manako the way we did.

people overestimate what can happen in a year and underestimate what can happen in 2, 5 or 10 over the last two weeks I had confirmations at so many levels that we were playing the right playbook at the right place and the right time, with the right team. 1. IBC Amsterdam and NRF Paris confirmed the need for a real open vision AI alternative (and made us proud of our models’ quality and efficiency). 2. the current petrol station rollout confirmed that Manako was plug-and-play and ready to be installed at the speed of light (5 mins process from box opening to live deployment) in june I didn’t know any of that. did you? maybe they were right after all , maybe it all happens at once