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 3

Nov 11Wed41–60
  1. 钛媒体68

    China burns 55% of its tokens on short dramas, while the US burns 58% on code: The flow of tokens reveals two different 'ways of life' for AI.

    The claim that "China uses 55% of tokens for short dramas and the US uses 58% for code" lacks an authoritative source upon verification, but the structural difference between Chinese and American AI applications—concentrated in content generation and code development respectively—does exist. Tokens are not equivalent to computing power. China's token consumption in short dramas is chain-based trial and error, while the US's token consumption in programming is single-point embedding into workflows. The real difference lies in who pays for the tokens.

    Why it matters: The claim that China uses 55% of tokens for short dramas and the US uses 58% for code lacks an authoritative source, but the structural difference between Chinese and American AI applications does exist.

  2. 中广互联 · 短剧74

    Central and satellite TV stations are rolling out a wave of new AI-simulated human short dramas: a transformation from "small screen" to "big screen"

    Central and satellite TV stations are rolling out a wave of new AI-simulated human short dramas, moving from pilot projects to regular scheduling, with themes shifting toward mainstream values such as revolutionary history and intangible cultural heritage. According to China Audio-Visual Big Data, about 128,000 micro short dramas were launched nationwide in the first quarter of 2026, of which AI micro short dramas accounted for over 95%; Hunan TV's "The Later Journey to the West" cost about 900,000 yuan per episode, only one-tenth of a traditional drama series of the same scale. Starting September 1, the "Measures for the Development and Management of Micro Short Dramas" officially took effect, requiring AI micro short dramas to add a notice label in a prominent position in each episode.

    Why it matters: AI short drama scheduling on central and satellite TV is moving from pilot cases to regular practice, with themes shifting toward mainstream values such as revolutionary heritage and intangible cultural heritage, offering trend references for the industry.

  3. DataEye短剧观察 · 澎湃商业与内容趋势73

    The '持械入宋' team: 10 people created a hit with 4 billion views, warning caution in entering the AI short drama industry

    The '持械入宋' team was interviewed, reminding ordinary people to be cautious about entering the AI short drama industry, saying that complete beginners face high difficulty and need long-term passion and computing power investment. The 10-person team produced a hit with over 4 billion total views, with computing power investment for the entire series exceeding 1 million yuan. DataEye data shows that in the first half of 2026, the rate of AI dramas on Douyin exceeding 100 million views was only 0.48%, rising to 0.95% in October alone.

    Why it matters: The 10-person team reviewed the computing power cost of the 4 billion view hit and the industry's rate of exceeding 100 million views, reminding that entering the industry requires long-term investment and trial-and-error accumulation, rather than blindly following trends to copy hits.

  4. 虎嗅 · AI影视与短剧产业81

    Short Dramas Run Rampant: Three Reshuffles

    The short drama industry has undergone three rewrites of pricing rules—from traffic-based content pricing, to platforms taking over the payer role, to new bills in the AI era—and the market size is expected to exceed 121 billion yuan in 2026. The rise of the Hongguo free model has rewritten revenue-sharing methods, and AI technology has lowered production barriers. However, there is a mismatch between capacity expansion and revenue opportunities: in the first half of 2026, only 1.3% of AI dramas crossed the 50 million viewership break-even line. As platforms tighten guaranteed payments and live-action drama capacity shrinks, individual creators and AI-native companies have become new variables, and the industry has entered an era of 'content inflation'.

    Why it matters: The three rewrites of pricing rules in the short drama industry—from traffic, to platforms, to AI—reveal a structural contradiction between capacity expansion and revenue opportunities, providing practitioners with a framework for recalculating costs and benefits.

  5. Reddit StableDiffusion · 创作者与工作流分享75

    PAN 2 released: open-source 29B video generation model

    PAN 2 released, an open-source 29B latent diffusion model supporting video generation up to 20 seconds, 720p, 24fps, under Apache 2.0 license. The publisher also released training code and provides integrations with Diffusers, FastVideo, and vLLM. The model enables an LLM to generate dense text states/scripts, timing events and actions, with camera nested under events but decoupled from actions.

    Why it matters: Open-source 29B video generation model and training code released, allowing creators to evaluate its 20-second 720p generation capability and LLM-driven script-level controllability.

  6. Recraft 官方博客68

    Gemini Omni Flash 1.2 launches on Recraft Studio, supporting 15-second video generation.

    Google released the Gemini Omni Flash 1.2 video model on November 5, 2026, and it is now available on Recraft Studio. Compared to 1.1, the single generation duration has increased from 10 seconds to 15 seconds, consistency and lip-sync are clearer, and Gemini API prices have dropped by about 30%.

    Why it matters: The new version increases the single generation duration from 10 seconds to 15 seconds and improves cross-segment consistency and lip-sync, making it suitable for short ad scenarios that need a complete mini-story.

Nov 10Tue
  1. No Film School71

    Adobe MAX 2026: New features for filmmakers and video editors

    Adobe MAX 2026 opens, and Premiere receives updates including the official availability of Color Mode and the Generative Media Tool. The Generative Media Tool brings video, sound effects, soundscapes, and music generation directly into the timeline, and supports Firefly as well as partner models from Google, Kling, Runway, Luma, and others.

    Why it matters: Within the Premiere timeline, third-party models such as Kling and Runway can be directly invoked to generate missing footage, further integrating video editing and AI generation workflows.

  2. Creative Bloq70

    Adobe Premiere Pro will introduce AI agents to handle tedious editing work.

    Adobe is adding Creative Agent to Premiere Pro, which can automatically apply effects, track masked subjects, and adjust color and exposure across an entire sequence. The new Generative Media Tool can generate video, sound effects, and music directly within the timeline, and supports integration with third-party models such as Kling; Generative Remove can remove objects from the frame, and Color Mode provides a color grading workspace.

    Why it matters: Adobe is integrating third-party models such as Kling into the Premiere timeline, allowing editors to directly generate assets, but the copyright boundaries of third-party models are more ambiguous than Firefly.

  3. 娱乐独角兽 · 搜狐公开专栏71

    "Armed Entry into Song" Becomes a Hit: An Amateur Troupe Garners 12 Billion Views — Are Historical Themes a New Gold Mine for AI?

    "Armed Entry into Song" has surpassed 12 billion views across all platforms within two months of its release, with 14.81 million in-app favorites on Hongguo. This AI short drama, produced by a six-person creative team with no prior film or TV experience, achieved a breakout through long-form narrative and historical research. The article analyzes its viral logic and argues that historical themes are becoming a new gold mine for AI short dramas, but hits are hard to replicate; topic selection judgment, persistence, and long-term operation are the real barriers.

    Why it matters: This review breaks down the advantages of long-form narrative and historical research in "Armed Entry into Song", providing a reference path for AI short drama creators to break out.

  4. Reddit comfyui · 创作与工作流分享67

    ComfyUI Tutorial: Using RelayVSR and DLSS5 to Fix Detail Loss in MiniMax H3 Video Upscaling

    ComfyUI tutorial uses RelayVSR and DLSS5 nodes to fix detail loss in MiniMax H3 video upscaling. Under 6GB VRAM, it can upscale a 0.7 MP video to 2K in about 7 minutes. Compared to MiniMax latent space upscaling which takes about 45 minutes, this method significantly reduces upscaling time.

    Why it matters: The RelayVSR workflow shared by the author can upscale MiniMax H3 video to 2K under 6GB VRAM in about 7 minutes, providing a portable upscaling solution for users with low VRAM.

  5. 钛媒体72

    Sell your face for 300 yuan, and the next second you become the protagonist of an AI short drama.

    Real-person portrait licensing is becoming a booming business. For 300 yuan, you can license your face to AI to generate a digital human, used for AI short dramas and product-promotion short videos. Overseas, Twinnin has about 2,000 users who have listed digital portraits; domestic platforms like 新爪 and ActID have launched matchmaking services. But once AI has seen a face, it is hard to forget, and the licensor cannot track where the face goes.

    Why it matters: Real-person portrait licensing has become a new business for AI short dramas, but after licensing, the face may enter a training set and become untraceable. The cost of rights enforcement is high, and the risk is borne by the licensor.

Nov 9Mon
  1. 电影界 · 短剧经营与内容趋势70

    In 2026, the micro-drama market size is expected to surpass 121 billion yuan, with AI dramas accounting for about 30%, approximately 40 billion yuan.

    In 2026, the domestic micro-drama market size is expected to exceed 121 billion yuan, a year-on-year increase of 21%, of which the market size of AI dramas and animated dramas is expected to reach 40 billion yuan, accounting for about 33.1% of the overall market. The National Radio and Television Administration has launched a two-month special governance campaign for micro-dramas starting from November 1, with AI micro-dramas and algorithm distribution mechanisms being systematically included in the special governance framework for the first time. The number of active users of overseas AI animated dramas has exceeded 150 million, and Chinese companies occupy nine of the top ten overseas short drama paid applications.

    Why it matters: AI dramas account for about 30% of the micro-drama market, and regulatory forward-shifting and overseas expansion are reshaping the industry's growth logic.

  2. DataEye短剧观察 · 澎湃商业与内容趋势74

    TikTok short dramas' September revenue share surpasses $32 million, with increased support for AI comic dramas.

    In September 2026, TikTok short dramas' total revenue share exceeded $32 million, with local dramas' monthly revenue share surpassing $25 million, and AI comic dramas' revenue share growing 200% month-over-month. The platform raised the incentive coefficient for local AI comic dramas this month to align with live-action dramas. In September, 59 AI short dramas received hit incentives totaling $1.16 million. Cumulative script licensing revenue share exceeded $1.2 million, with 39 IP licensing deals each generating over $10,000 in revenue share.

    Why it matters: TikTok short dramas' September revenue share surpasses $32 million, AI comic dramas grow 200% month-over-month and receive platform coefficient alignment support, providing a reference for the commercialization of AI short dramas going overseas.

  3. InvokeAI 官方发布75

    InvokeAI 7.0.0 beta 1 released

    InvokeAI 7.0.0 enters beta 1, its largest version update. It adds projects, a single workbench, and video panel support for MiniMax H3, LTX-2.5, and Wan 2.2. Official warning: v7 database migration is irreversible; back up before upgrading.

    Why it matters: InvokeAI 7.0.0 enters beta, adding projects, a video panel, and multi-model support, giving creators a basis to evaluate the upgrade.

  4. Morketing · 出海商业洞察71

    On the eve of the AI interactive drama boom, will the short drama and game businesses be remade?

    Morketing published an analysis of the form and business model of AI interactive dramas. DRAMAGO has launched 14 AI series, 10 of which are interactive content; the completion rate for the first three episodes is nearly twice that of live-action short dramas on the platform, and the conversion rate at the first payment point is nearly 30 times higher; Castloop has listed over a hundred original works, and StoReel has raised $34 million in funding.

    Why it matters: AI interactive dramas are at the intersection of short dramas and games. Small and medium-sized teams have already validated the possibility of interactive payment and retention with low-cost content.

  5. 中广互联 · 短剧73

    AI short drama 'Armed Entry into Song' had an investment of over 1 million yuan and a return of about 10 million yuan.

    AI short drama 'Armed Entry into Song' had computing power investment of over 1 million yuan and commercial returns of about 10 million yuan, with over 6.6 billion views across the internet, becoming a hit in the AI short drama track. The drama was produced by a 6-person non-professional team using Dreamina 2.5, with no paid traffic, and relied on content completeness to receive spontaneous recommendations from celebrities. However, industry data shows that only 1 in 77 new dramas can recoup costs, and over 90% of teams are losing money; the creators urge people not to blindly follow the trend.

    Why it matters: 'Armed Entry into Song' used over 1 million yuan in computing power investment to obtain about 10 million yuan in returns, but in the industry, only 1 in 77 recoups costs, providing practitioners with a real reference where hits and intense competition coexist.

Nov 8Sun
  1. Stable Diffusion Art 教程68

    Generating 360-degree orbit bullet-time videos with ComfyUI and MiniMax H3

    Stable Diffusion Art provides a set of ComfyUI workflows that chain three MiniMax H3 videos together to generate a 360-degree orbit bullet-time shot from a single photo.

    Why it matters: Chaining three MiniMax H3 videos into one ComfyUI workflow can generate a bullet-time shot with a 360-degree orbit from a single photo.

  2. 文娱价值官 · 搜狐公开专栏79

    Interview with the creators of 'Armed Entry into Song': How a grassroots team with no formal training and no paid promotion made a 12-billion-view hit

    Wenyu Value Officer interviewed the creative team 'Lao Shiyi' of 'Armed Entry into Song'. The team, consisting of 6 non-professional members, with no paid promotion and only about 1 million yuan in computing power costs, produced a hit short drama with 12 billion views that topped the Hongguo AI list. The team revealed they mainly used the Seedance2.5 model; script polishing and gacha are the biggest time costs, with 30 people collaborating online, none with a film/TV academic background.

    Why it matters: The creative team publicly shared the creation process and cost structure of a non-professional, no-paid-promotion team, providing a hit methodology that AI short drama creators can draw on.

  3. Variety 产业新闻68

    Korean AI production pioneer warns at Taiwan Creative Content Festival: AI could make film production costs exceed those of the 35mm era

    Korean AI production pioneer warned at the Taiwan Creative Content Festival (TCCF) that AI could make film production costs exceed those of the 35mm era, due to token fees and piecemeal revisions.

    Why it matters: Korean AI production pioneer warns that token fees and piecemeal revisions could make AI feature film costs exceed those of the 35mm era, providing a cost reference for producers to evaluate AI efficiency.

  4. Reddit comfyui · 创作与工作流分享71

    Remiqora Video launches: generate a complete video with sound on a local PC from a single sentence or a song

    Remiqora Video releases a free and open-source video studio based on ComfyUI. Enter a sentence or a song and it generates a complete video with sound on a local PC. Supports 15 seconds to 10 minutes, landscape/portrait/square aspect ratios. Songs can have lyrics automatically recognized and be cut to the beat. Characters and scenes remain consistent across shots. Individual scenes can be re-shot separately, and bad shots are automatically detected and re-shot.

    Why it matters: Free, open-source, and runs locally. It wraps a ComfyUI workflow into a complete application that produces a video from a single sentence. Supports single-scene re-shooting and character consistency, and can evaluate local video production efficiency.