This episode pairs an AI video news rundown with a hands-on tutorial on chaining keyframes into continuous multi-shot scenes. ## News (window June 15-19, 2026) - **Grok Imagine Video 1.5 hits GA** on 2026-06-16 across the Imagine API, web, and mobile, with a high-speed Fast variant the same week ([explainx](https://explainx.ai/blog/grok-imagine-video-1-5-xai-release-2026), [gagadget](https://gagadget.com/en/715382-grok-imagine-video-15-hits-number-one-and-its-86-cheaper-than-sora/)). Image-to-video with native synchronized audio; reportedly 720p, 24fps, 6-second clips; Fast renders 6s/720p in about 25 seconds ([GIGAZINE](https://gigazine.net/gsc_news/en/20260618-grok-imagine-video-1-5/), [winbuzzer](https://winbuzzer.com/2026/06/18/xai-speeds-up-grok-imagine-video-15-for-ai-video-creators-xcxwbn/)). Verify-grade pricing: $0.080/sec, 60 req/min ([xAI Docs](https://docs.x.ai/developers/models/grok-imagine-video-1.5-preview)); press frames 720p at $4.20/min, ~86% below legacy Sora 2 Pro. - **ByteDance Seedance 2.0 Mini** lands on Dreamina mid-June, broader API ~2026-06-22; ~2x faster than Seedance 2.0 Fast at ~$0.073/sec, keeps the 12-reference multimodal system ([Pexo](https://pexo.ai/blog/what-is-seedance-2-0-mini-1033), [Atlas Cloud](https://www.atlascloud.ai/blog/ai-updates/seedance-2.0-mini-overview)). - **Runway Studio Trim** (2026-06-18): trim, stitch, reorder, export inside Studio on all tiers ([Releasebot](https://releasebot.io/updates/runwayai)). - **Arena Elo**: Seedance 2.0 leads both I2V boards; Grok 1.5 sits #2-3, not the "#1" press claims ([AA I2V](https://artificialanalysis.ai/video/leaderboard/image-to-video)). ## Tutorial: Keyframe Chaining Why clips cap at 5-10 sec, the last-frame to start-frame workflow, the frozen-seam problem ([CineLOG](https://arxiv.org/pdf/2512.12209)), first/last-frame interpolation per model (Kling, Luma, Runway, Pika, Veo 3.1, Wan FLF2V, Seedance), native extend vs manual chaining, fighting temporal drift and color shift ([Knot Forcing](https://arxiv.org/pdf/2512.21734)), and a copyable end-to-end workflow with DaVinci Resolve and CapCut ([tensorpix](https://tensorpix.ai/blog/ai-video-continuity-consistent-scenes), [seedance-2ai](https://seedance-2ai.org/blog/ai-video-first-last-frame-guide)).