Qwen3.5 0.8B Real-Time Video Captioning on Mac Studio
Melvin Vivas · X video post · 2026-03-26 · Open on X
Topics: LLM Fundamentals, Industry Trends & Job Market · Level: intermediate
Summary
This reshares the same demo: Qwen3.5 0.8B (about 1GB) captions video locally on a Mac Studio M2 Ultra at under 1s per frame. It streams scene descriptions while the video plays.
Key points
- The roughly 1GB Qwen3.5 0.8B model handles real-time video captioning.
- Under 1s per frame; 269 frames from a 3m49s video.
- Runs fully locally on a Mac Studio M2 Ultra.
- It understands the scene at any point where you pause.
Resources mentioned
- Qwen 3.5 · tool · qwen.ai · free
Alibaba Qwen open model family with small on-device variants and toggleable reasoning.
Also in: Workshop AI: Building Apps with Cloud and Local Agents (GLM 5, Qwen 3.5) (Melvin Vivas on X · notes), Qwen3.5 0.8B Does Real-Time Local Video Captioning (Melvin Vivas on X · notes), Qwen 3.5 Is the Local Model That Holds Up in Claude Code (Melvin Vivas on X · notes), Free, Unlimited Agentic AI with Local Qwen 3.5 (Melvin Vivas on X · notes) and 7 more
Try this
- Build a streaming video description tool with a small local vision-language model.
More in LLM Fundamentals
- 1-bit Bonsai 8B Runs On-Device on iPhone at 40+ tok/s
- MiniMax 2.7 One-Shots a Linear Clone at 95% Lower Cost
- Qwen3.5 0.8B Does Real-Time Local Video Captioning
- How to Get a GLM-5-Turbo Rate Limit Increase for OpenClaw
- GLM-5-Turbo: A Fast GLM-5 Variant for Agents Like OpenClaw
- Real-Time Speech Transcription in the Browser with Voxtral and WebGPU