Skip to content

Creative Directions

Video Generation Latest

The front line of text-to-video and image-to-video: tracking model releases, capability boundaries, price changes, and transformations in film and television creation.

344 picksLast 30 days 240 itemsCollected 1,725 items

updated

Pick archive · page 7

Nov 2Mon121–140
  1. Stuubzzz · 视频工作流实测75

    MiniMax H3 Character LoRA Training Test: 29 Clips, 4000 Steps, Why Step 3000 Was Chosen

    The author trained a MiniMax H3 character LoRA using AI Toolkit on an RTX 5090, with 29 3-second clips, rank 16, 4000 steps, taking about 15 hours of GPU time, and ultimately chose the step 3000 checkpoint.

    Why it matters: The author tested training a MiniMax H3 character LoRA using AI Toolkit on an RTX 5090, and provides the specific settings of 29 clips and 4000 steps, along with 5 reusable rules for reference when training other video models.

  2. Stuubzzz · 视频工作流实测70

    MiniMax H3 RefMod and reference image testing: long-video facial consistency comparison

    The author's testing found that MiniMax H3's RefMod can improve long-video facial matching from 0.23 to 0.80; the 0.5 MP version scored 0.75 in the second segment compared to 0.25 without RefMod.

    Why it matters: Testing shows that MiniMax H3's RefMod can improve long-video facial matching from 0.23 to 0.80; the 0.5 MP version offers better value.

Nov 1Sun
  1. 电影界 · 短剧经营与内容趋势69

    The three major long-form video platforms are collectively entering the AI comic-drama agent space, as iQIYI, Tencent, and Mango bring the tool-layer arms race into their own ecosystems.

    In October 2026, the three major long-form video platforms iQIYI, Tencent, and Mango collectively launched AI comic-drama creation tools, respectively releasing Nadou Pro, WorkRally, and the Shanhai AIGC creation platform, internalizing the AI production chain into their own ecosystems.

    Why it matters: The three major long-form video platforms have internalized AI creation tools into their own ecosystems, potentially re-dividing the supply landscape for short dramas and comic dramas, with tools shifting from third-party plugins to platform infrastructure.

  2. 电影界 · 短剧经营与内容趋势68

    Chinese Online's first-three-quarter revenue was 833 million yuan, down 17.64% year-on-year; AI short drama cost reduction drove a 74.63% loss reduction.

    Chinese Online's revenue for the first three quarters of 2026 was 833 million yuan, down 17.64% year-on-year; net profit attributable to the parent company was a loss of 132 million yuan, a 74.63% reduction in loss year-on-year. The loss reduction mainly came from AI short drama cost reduction; operating costs for the first three quarters were 444 million yuan, down 32.91% year-on-year, with part of overseas short dramas converted to AI short dramas. The company terminated its H-share listing plan, received an inquiry letter from the Shenzhen Stock Exchange regarding its 2.833 billion yuan private placement, and net assets attributable to the parent company fell to 175 million yuan.

    Why it matters: The AI short drama cost reduction strategy is reflected in Chinese Online's third-quarter report as a 74.63% loss reduction, but the 17.64% revenue decline shows that the transition from stopping the bleeding to generating cash flow still takes time.

  3. Reddit StableDiffusion · 创作者与工作流分享77

    Author creates 46-second sci-fi trailer entirely locally on MacBook M5 Max and reveals complete pipeline

    Author completes 46-second sci-fi trailer 'THE LAST SIGNAL' entirely locally on a MacBook Pro M5 Max 128GB, with no cloud rendering and no API calls, taking about 14.5 hours (including about 2 hours of waiting).

    Why it matters: The author reveals the complete local pipeline from storyboard to final film and the actual time taken, allowing creators to assess the feasibility of replicating a similar workflow on a MacBook.

  4. 虎嗅 · AI影视与短剧产业80

    The author of the AI comic drama '万妖图录传' with 6 billion views earned less than 1 million yuan.

    The author of the AI comic drama '万妖图录传' with 6 billion views earned less than 1 million yuan across twelve seasons, and the first season with 2 million collections only earned 20,000 yuan. The article uses cases from 番茄, 红果, 阅文, 掌阅, and 中文在线 to sort out the revenue-sharing rules of the AI comic drama industry chain, pointing out that having many IPs does not necessarily mean pricing power, and having an outlet does not necessarily mean making money. 红果 controls the traffic outlet and revenue-sharing rules, but faces cost pressures such as declining ad unit prices, declining coin subsidies, competitors grabbing content, and stricter regulation.

    Why it matters: Through comparison of multiple company cases, it points out that in the AI comic drama industry chain, having many IPs does not equal pricing power, and having an outlet does not necessarily mean making money, providing a reference for content parties to evaluate platform choices.

  5. Reddit comfyui · 创作与工作流分享70

    The author used MiniMax H3 and the self-developed Video Builder in ComfyUI to create a Halloween music video.

    The author used the self-developed Video Builder to have Claude automatically complete lyrics, prompts, and the render queue in ComfyUI, producing a MiniMax H3 Halloween music video. The entire video has 19 scenes, 2:29, with 80 renders in total and about 12 GPU hours; the author manually handled story ideas, the Suno song, and Nano Banana scenes, while Claude automatically generated characters, planned shots, and checked lip sync.

    Why it matters: The author published the complete division of labor and 2-pass workflow for Claude-driven ComfyUI music video production, which can be reused in one's own video generation pipeline.

Oct 31Sat
  1. 漫剧出海 · 每日观察80

    ReelShort hands nearly 90% of its content to AI: supply is being rewritten, and channels are starting to tighten.

    ReelShort's AI content share rose from 10% in May to 88% in September, while live-action production remained stable at 10–30 titles per month; during the same period, TikTok for the first time listed art style as a separate governance category, Meta banned ByteDance-affiliated ads in seven countries, and the supply and distribution rules for short dramas going overseas were rewritten simultaneously.

    Why it matters: As AI content share and channel governance tighten in tandem, short dramas going overseas are moving from a wild-west phase to a professional stage, and practitioners need to reassess their content and distribution strategies.

  2. 传媒内参 · 搜狐公开专栏72

    The AI documentary race among iQiyi, Youku, Tencent Video, and Mango TV has begun.

    The four platforms—iQiyi, Youku, Tencent Video, and Mango TV—are accelerating their AI documentary strategies: Tencent Video has launched 11 works in its "AI Lingjing Season," iQiyi plans to release 100 AI documentaries by the end of the year, Mango TV's "Taiwan Island Chronicle" is airing on satellite TV, and Youku is using generative AI to restore classic old films. AI is transforming documentaries from "only subjects you can afford to shoot are worth making" into "subjects worth telling can all be seen."

    Why it matters: The four platforms take different paths—from scaled matrices and full-process AI to satellite benchmarks and old film restoration—offering a comparative assessment of the different strategies in the AI documentary track.

  3. 传媒内参 · 搜狐公开专栏74

    AI short-drama companies barely need entry-level roles, trending on hot search; behind the industry reshuffle are job upgrades and lowered thresholds.

    AI short-drama companies barely need entry-level roles, trending on hot search; traditional crews of 30–40 people have been compressed to 3–5, with basic execution roles taken over by AI tools. Demand for AI editors grew 179% year-on-year; the cost of premium AI live-action-style short dramas is about one-tenth that of live-action versions, and the entrepreneurial threshold has dropped from millions of yuan to tens of thousands. The industry is shifting from execution to creativity, entering an era of capability-based pricing.

    Why it matters: Behind the AI short-drama industry reshuffle are job upgrades and lowered entrepreneurial thresholds; ordinary creators can use this to assess entry paths and capability-pricing trends.

  4. 钛媒体65

    AI films are driving cinemas into a “dead end”.

    The first fully AI-produced theatrical film 'Sanxingdui: Future Past' underperformed at the box office after release; as of October 26, cumulative box office was 1.216 million yuan, with AI predicting a total of 2.609 million yuan. The production team generated over 1.2 million base images and conducted hundreds of full-film iterations, but audiences reported stiff character movements and abrupt shot transitions. The article argues that AI films still face multiple contradictions in production quality, audience acceptance, and theatrical business models.

    Why it matters: The first AI theatrical film 'Sanxingdui: Future Past' fell far short of box-office expectations, providing a reference for market acceptance and commercial returns as AI films enter theaters.

  5. note · cocoredo_2509 · 图像视频创作分享68

    Pika H3 Flex: Import, Quality, and Cost Verification for Video Creators

    The author verifies Pika H3 Flex from a production-site perspective, pointing out that official performance cannot be directly equated with actual usability; character consistency, audio track separation, and 4K upscaling need to be evaluated separately. H3 Flex is based on MiniMax H3, supports up to 12 reference materials and 5 types of audio track separation, and after generating at 480p/768p can be upscaled to 1080p/2K/4K, but 4K is upscaled rather than natively generated.

    Why it matters: On-site verification of Pika H3 Flex shows that character consistency, audio track separation, and 4K upscaling need to be evaluated separately; the author's three short tests and recording method can be directly used for other model adoption decisions.

  6. IT之家68

    Kuaishou's Kling AI completes approximately 19.8 billion yuan capital increase

    Kuaishou's Kling AI completes approximately 19.8 billion yuan capital increase closing, about 645 million yuan less than the original cap. Two investors failed to complete payment within the deadline, Beijing Kling terminated their capital increase, and the remaining shareholders all completed their capital contributions, including Shanghai Guofang Digital Technology, National AI Industry Investment Fund, Alibaba Cloud, etc.

    Why it matters: Kling's independent financing completed closing, with approximately 19.8 billion yuan capital increase landed, about 645 million yuan less than the original cap, due to two investors failing to complete payment.

  7. comfyorg · YouTube 创作视频65

    Pika H3 Flex is now integrated into ComfyUI.

    Pika H3 Flex is now integrated into ComfyUI, supporting T2V, R2V, and FL2VA workflows, allowing you to go from draft to 4K final film within a single node graph. The model renders a 15-second 768p clip with audio in less than 10 seconds, supports up to 12 reference inputs and 5 independent audio tracks, and ranks 7th on the Design Arena text-to-video leaderboard.

    Why it matters: After Pika H3 Flex is integrated into ComfyUI, a 15-second 768p clip with audio renders in less than 10 seconds, supports 12 reference inputs and 5 audio tracks, making it suitable for batch iteration.

  8. Curious Refuge 创作教程69

    How to Use AI to Relight a Video (Step-by-Step Tutorial)

    Curious Refuge published a tutorial demonstrating how to use AI to relight videos, and tested the workflow's effectiveness on live-action footage and gray-box studio scenes. The workflow first exports still frames and relights them using an image model (Nano Banana 2.1 or GPT Image 2.5), then uses the relit images as references to apply to the original video. Testing found limitations such as exposure shifts, banding, and resolution loss, with output resolution lower than the 4K source.

    Why it matters: This workflow uses static-image relighting as a reference for the video model, allowing creators to evaluate the boundaries and limitations of AI relighting in post-production.

  9. Higgsfield 官方博客76

    Higgsfield's real-world test of 2026 AI video costs: a 10-second finished video ranges from $1.35 to $7.85

    Higgsfield's test shows that a single generation of a 10-second AI video costs about $7.85 (160.125 credits) on Seedance 2.5 at 1080p, and about $1.35 on Kling 4.0 at 720p. The actual cost must also account for retakes, image preparation, and post-production.

    Why it matters: Provides a method and budget formula for calculating the true cost of AI videos under per-generation pricing, allowing creators to plan retake counts and monthly output accordingly.

  10. Reddit comfyui · 创作与工作流分享67

    Same shot, same camera, different versions: a ComfyUI app reshoots your footage

    An open-source ComfyUI app can reshoot footage while preserving camera, composition, and architecture. After uploading a clip, choose a scene (future, abandoned, flood, golden hour, cyberpunk, anime), approve a new first frame, and the app regenerates the entire shot. Using only core nodes and open weights (Qwen-Image-Edit + LTX 2.5 + SAM 3), the demo and repository are public.

    Why it matters: An open-source workflow uses core nodes and open weights to achieve shot reshoot, preserving camera, composition, and architecture, allowing creators to migrate this to their own production pipeline.

Oct 30Fri
  1. Reddit StableDiffusion · 创作者与工作流分享67

    PxTicks open-sources Minimax H3 motion control workflow and companion applications

    PxTicks has open-sourced the Minimax H3 motion control workflow, which allows controlling motion by dragging objects, and comes with background separation, motion path, and mask generation tools that can be directly used with the ComfyUI interface.

    Why it matters: The author has open-sourced the Minimax H3 motion control workflow, which allows controlling motion by dragging objects; creators can refer to it to achieve similar effects.

  2. Reddit comfyui · 创作与工作流分享71

    MarcoStudio open-sources 13 image and video tools, running locally on ComfyUI

    MarcoStudio open-sources 13 image and video tools, running locally on ComfyUI, with all workflows and models under the MIT license.

    Why it matters: The open-source ComfyUI toolset provides 13 image and video generation features, all running on local GPUs, allowing creators to evaluate localized creative workflows.

  3. MindStudio Blog · 创作分享70

    3D to AI Video Workflow: How to Maintain AI Character Consistency

    The 3D-to-AI video workflow uses 3D models as references to solve the problem of AI video characters drifting as camera angles change, involving model generation tools such as Trippo 3D, Hunyuan 3D, and Meshi V7. Among them, Meshi V7 has flexible export formats, Cozy Clay allows manipulating models in the browser, and Simulon can place characters into real environments to generate videos. However, the generated models are not rigged with skeletons by default, and animation still requires additional rigging steps.

    Why it matters: A workflow using 3D reference models to solve AI video character drift, and comparing the applicable scenarios of tools such as Trippo 3D, Hunyuan 3D, and Meshi V7.