Creatify post-trains MiniMax H3 into ad video model Boreal-H3, lifting brief success from 28% to 50%
testingcatalog · x · 2026-10-01
Creatify has launched Boreal-H3, a video model built for ad production, along with an Ad Agent powered by Claude Opus 5.5.
Boreal-H3 was post-trained on MiniMax H3 using real ad projects. Reported metrics on the same task set: brief success rate rose from 28% to 50%, identity match from 83% to 94%, and failure severity dropped 71%.
The model targets ad-specific requirements: products keep their shape, labels and details as the camera moves; generated virtual creators handle UGC, complex spoken lines, and two-person dialogue; camera movements and keyframes can be user-directed. Creatify says this isn't a one-off SFT/LoRA fine-tune but a closed-loop system: a human-calibrated evaluator diagnoses failures and guides the next intervention—targeted data collection, RL, or inference optimization.
Related event: Creatify's MiniMax H3-based ad video model lifts brief success to 50%(2 posts)→
More from Multimodal
- Testing Qwen Image 2.1 and Minimax H3 on anime-to-photo conversion and film photo restoration — Ant_6431 · 2026-11-16
- Second take of a 'Cubee run' video generated with Minimax H3 — JahJedi · 2026-11-16
- Non-artist builds 14-chapter hand-painted storybook game with Fable 5.2 in 4 hours — jradoff · 2026-11-16
- Subject LoRAs on Minimax H3 Image learn poor likenesses while identical datasets excel on Qwen 2.1 — Ad_Organic · 2026-11-16
- Seedance-generated anime train scene stuns Reddit with cinematic quality — Ok-Solution-7889 · 2026-11-16
- AI-generated 'Men in Black' reimagined in 1977 goes viral — techhalla · 2026-11-16