New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS
Open-source tools like Joy AI Video Edit and Tencent Scope give AI filmmakers real-time prompt-driven editing and fine-grained camera motion control locally.
What this lesson covers
This AI news roundup breaks down several model releases including Joy AI Video Edit for open-source video editing and Tencent's Scope for direct camera trajectory control. It also covers frontier LLM updates such as DeepSeek V4 Pro, Grok 4.6, and Gemini 3.7 Flash.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 00:49 ↗
Joy AI Video Edit performs open-source, prompt-driven video modifications such as outfit changes and background removal with low latency.
- 01:48 ↗
Joy AI Video Edit uses a 16-billion-parameter multimodal diffusion transformer architecture capable of generating 720p video at over 30 fps.
- 02:30 ↗
Tencent's Scope framework provides precise camera trajectory control by combining an input image with a specified camera motion path.
- 03:36 ↗
DeepSeek V4 Pro August 13th introduces D-Spark speculative decoding across a 1.7-trillion-parameter MoE architecture at low API costs.
- 06:27 ↗
Grok 4.6 achieves frontier-level benchmark scores across professional tasks while maintaining competitive token pricing.
Workflow outlined in the video
- Download the Apache 2-licensed Joy AI Video Edit weights to test real-time prompt-based video editing locally.
- Implement Tencent Scope in your image-to-video workflow to enforce specific camera trajectories like dollies, cranes, and S-curves.
- Set up DeepSeek Harness to test open-source agentic scripting and workflow automation.
Before you use this workflow
These notes describe the source video at its publication date. Model access, pricing, connectors and interfaces may have changed. Check the original source and the provider’s current documentation before installing an add-on, connecting an account or spending credits.
ReelStack has not independently tested this workflow. Preview one representative shot and check motion, continuity and output quality before applying it to a full production. No result, cost saving or model capability is guaranteed.
How to interpret the numbers
155,956 views captured 27 September 2026. ReelStack recommendation score: 37/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 45% performance: 1.02× views vs 8 other channel videos observed at a similar age. View-evidence factor 99% (views / (views + 1,000))
- 25% freshness: 12/100 with a 14-day half-life
- 20% momentum: 193 views/day vs 822 channel baseline (6 comparable uploads), with the same view-evidence factor
- 10% engagement: 76/100; likes + 4× comments, smoothed with a 500-view neutral prior