SenseTime Open-Sources SenseNova 6.8 Code2Video: Prompt-to-Video With Editable Code
FellMentKE · x · 2026-11-06
SenseTime's SenseNova 6.8 Code2Video is the first open-source model for code-driven video creation: it builds and renders videos by generating executable code, with output including editable source for visuals, animation, timing, and scenes that developers can inspect and modify.
It can also recreate reference videos and refine them through conversation. The release includes model weights, Skills, prompts, source code, and creation traces. The 90-second launch film was itself entirely generated by the model. It currently focuses on short videos implemented fully in code, with promising visual composition and motion choreography. The GitHub repo has 528 stars.
More from Multimodal
- AI-made video hits 1M views in 4 hours, but creator insists every edit was human-made — WesEklund · 2026-11-06
- Blender × Claude Opus 5.5 demo shows near-autonomous 3D modeling quality — bilawalsidhu · 2026-11-06
- Turn NVIDIA PiD into a universal 4x upscaler via FLUX.2 VAE re-encoding — TimeTruth2490 · 2026-11-06
- Iterative AI ship design: from target shape to structured hull in about 10 minutes — Promptmethus · 2026-11-06
- Korean team's AI-made K-pop MV "Higher" looks industry-grade — Beginning_Victory597 · 2026-11-06
- RelayVSR is the new king of video upscalers, 20x faster than SeedVR2 — CeFurkan · 2026-11-06