Weekly cut notes: dated shipping that changes a cut. Credit the writer. Links only. https://ai-fieldnotes.com/news/ 2026-10-19T12:00:00Z https://ai-fieldnotes.com/news/#topview-3d-builder-2026-10-15 2026-10-15T12:00:00Z

Credit: Topview. Official Topview release (15 Oct 2026): 3D Builder stages an editable 3D scene from a description or reference image, so you can place characters and props, set the camera and its move, and render reference frames before spending a video generation. A local open-source route needs no account; the hosted build is $0.10 a scene. Previs and blocking, not finished footage.

Topview https://ai-fieldnotes.com/news/#luma-shot-coverage-2026-10-15 2026-10-15T12:00:00Z

Credit: Luma. Official Luma launch (15 Oct 2026): Shot Coverage turns one video take into extra moving camera angles you can cut with, so a single shot can be edited like covered footage. In the Luma app; check credits on Luma's plans page before a big run.

Luma https://ai-fieldnotes.com/news/#kling-4-0 2026-10-15T12:00:00Z

Credit: Kling AI / Kuaishou. Official Kling release notes (15 Oct 2026): Kling 4.0 is now live for everyone on kling.ai. Native 3 to 30 second generation in one pass, up to 10 keyframes along the timeline, and Omni Reference with up to 15 image, video or voice references; also reachable through Kling MCP and CLI. Native 4K and 1080p 10-bit HDR output and API access for developers are still marked coming soon. Check your plan on kling.ai.

Kling AI / Kuaishou https://ai-fieldnotes.com/news/#heygen-voice-2026-10-14 2026-10-14T12:00:00Z

Credit: HeyGen. Official HeyGen research post (14 Oct 2026): HeyGen Voice clones a voice from a short recording and shapes delivery and room acoustics to match a target scene image, including mic distance. Ranked first on Artificial Analysis's controlled-voice leaderboard (9 Oct 2026 snapshot). Useful for dialogue and voice-over; built mainly around HeyGen's talking-avatar videos.

HeyGen https://ai-fieldnotes.com/news/#higgsfield-katana-2026-10-12 2026-10-12T12:00:00Z

Credit: Higgsfield. Official Higgsfield editorial (12 Oct 2026): Katana is an editing skill for Claude via Higgsfield MCP. Give it a reference edit (upload or library preset) plus up to five photos of characters, products or clothing; it rebuilds the shot structure, pacing and cuts with your assets, then iterates in the same chat. Uses your existing Higgsfield plan credits (no separate MCP bill). Needs Claude with the Higgsfield connector and an active Higgsfield subscription. Product door and setup on higgsfield.ai/katana.

Higgsfield https://ai-fieldnotes.com/news/#vidu-q4-preview-2026-10-07 2026-10-07T12:00:00Z

Credit: Shengshu Technology / Vidu. Official launch (7 Oct 2026): Vidu Q4 Preview is the first public preview of the next Vidu model. Image-to-video and reference-to-video with native audio: up to 15 image refs and 3 voice refs, 3 to 16s clips with multi-shot camera switching, 2K/4K out with 10-bit colour. Pitched at short drama, ads and film teams. Live on the Vidu web app and API (viduq4-preview) ahead of the full Q4, so check current pricing and limits on vidu.com.

Shengshu Technology / Vidu https://ai-fieldnotes.com/news/#nano-banana-2-1-deepmind 2026-10-06T12:00:00Z

Credit: Google DeepMind. DeepMind model card (6 Oct 2026): Nano Banana 2.1 upgrades the Gemini image line on visual design, mask-based editing and subject consistency. Stills gen and conversational edit; listed for Gemini, AI Studio, the Gemini API, Search AI Mode, Ads, Stitch and Google Flow. Useful for boards and key art next to a Flow cut. Check live access on those surfaces; API id gemini-nano-banana-2.1 on ai.google.dev.

Google DeepMind https://ai-fieldnotes.com/news/#utopai-x-pai-2026-09-30 2026-09-30T12:00:00Z

Credit: Utopai Studios. Official Utopai Studios note (30 Sep 2026): Utopai X, a text-to-video model with audio built inside a working film and TV studio, now runs in the all-new PAI production workspace. Script breakdown, references and every take stay tied to the shot, so you compare versions instead of rebuilding the brief. The studio cites No. 2 on Artificial Analysis' text-to-video with audio board (29 Sep). Paid plans; footage exports to Premiere or Resolve. Check the current plan on utopaistudios.com.

Utopai Studios https://ai-fieldnotes.com/news/#veo-3-1-deepmind 2026-09-30T12:00:00Z

Credit: Google DeepMind. DeepMind’s current Veo page. Access sits in Flow, Gemini and Google’s other first-party tools; check the current plan there, not on a reseller.

Google DeepMind https://ai-fieldnotes.com/news/#grok-imagine-access 2026-09-30T12:00:00Z

Credit: xAI. Official Grok page: Imagine does stills and video from text or a reference photo, with follow-up edits in the same thread. Video listed up to 15 seconds. Open it on grok.com.

xAI https://ai-fieldnotes.com/news/#eleven-v4 2026-09-28T12:00:00Z

Credit: Mati Staniszewski & Piotr Dabkowski, ElevenLabs. Official launch: Eleven v4 and low-latency v4 Turbo. More emotional range, stronger voice cloning, 90+ languages. Live in ElevenAgents, ElevenCreative and the API.

Mati Staniszewski & Piotr Dabkowski, ElevenLabs https://ai-fieldnotes.com/news/#sora-api-shutdown-2026-09-24 2026-09-24T12:00:00Z

Credit: OpenAI. Official help note and API deprecations: consumer Sora already closed; Videos API and sora-2 / sora-2-pro shut on 24 Sep 2026 with no named replacement. Check exports on OpenAI before you plan around it.

OpenAI https://ai-fieldnotes.com/news/#runway-davinci-resolve-2026-09-23 2026-09-23T12:00:00Z

Credit: Runway. Official Runway note (23 Sep 2026): free plugin for DaVinci Resolve Studio 19+ puts Runway generation and Aleph restyle on the timeline. Results import to the Media Pool without leaving the edit. Paid Runway plan credits. Check download and Studio requirement on runway.com. Pipeline chaining lives on Workflows (Runway Workflows).

Runway https://ai-fieldnotes.com/news/#gen4-formula-e 2026-09-15T12:00:00Z

Credit: GEN4 / Formula E. Up to 60 seconds, made with Google AI models and the GEN4 asset kit. Premiere listed at BFI IMAX, 22 Oct 2026. Check the call on their page.

GEN4 / Formula E https://ai-fieldnotes.com/news/#suno-v6 2026-09-09T12:00:00Z

Credit: Mikey Shulman, Suno. Official post: Suno introduces v6, v6-wild and v6-mini, with more control across editing, mashups, sampling, lyric updates, and text, audio, image and video inputs.

Mikey Shulman, Suno