Skip to content
Date and source

Dates follow the news timeline in Beijing time and include both endpoints.

TodayNov 19Thu1 item
  1. Google DeepMind68

    Genie 4 is now available.

    Genie 4 is now available to Google AI Ultra subscribers, supporting real-time world editing during sessions and real-time native audio. Genie 3 has been updated in sync, with single generations up to 3 minutes long, supporting 360-degree panorama seeds and real-time world events; new Project Genie agent and community gallery have been added.

    Why it matters: Genie 4 is available to Google AI Ultra subscribers, supporting real-time world editing during sessions and real-time native audio, allowing creators to evaluate its interactive world generation, sharing, and re-creation capabilities.

Nov 18Wed
  1. 镜像娱乐 · 搜狐公开专栏51

    Paradigm Intelligence releases video large model Phai-M1, targeting live-action film and television scenes and short drama production.

    Paradigm Intelligence releases video generation model Phai-M1, optimized for live-action film and television scenes, solving the problems of character distortion and audio-video desynchronization, and supporting synchronized audio-video generation. In actual testing, generating a 15-second 720P video takes about 3 minutes. The model completed full-process training on Huawei's Ascend platform, and the company plans to open-source the weights soon.

  2. 爱范儿65

    Jimeng Seedance 2.5 launches a 4K version, supporting up to 30 seconds of single generation and six aspect ratios.

    Jimeng Seedance 2.5 launches a 4K version, supporting up to 30 seconds of single generation, full-modal reference, video editing and extension, and adopts 10-bit BT.709 color format and six aspect ratio outputs.

    Why it matters: The author tested Seedance 2.5 4K in four scenarios: brand TVC, e-commerce, film/TV previsualization, and large-screen display, demonstrating the practical value of 4K quality for commercial production workflows.

  3. Monster社区 · 中文ComfyUI实测与工作流56

    How to choose Seedance 2.5 Flash: comparing visuals, narrative, and cost

    The author conducted hands-on comparisons between Seedance 2.5 Flash and the original version, recommending that rework costs be factored into the selection; projects with strong narrative elements tend to favor the original. Testing found that Flash offers better visual texture, but complex storytelling tasks are prone to motion blur, blurring, or logic issues. This article did not reproduce the 480P super-resolution operation mentioned in the article.

  4. 可灵官方博客65

    Kling releases Kling Image 4.0, focusing on realism and native 8K generation.

    Kling has released its next-generation image generation model, Kling Image 4.0, now available on Kling Creative Studio and open to enterprises and developers via API. The model is based on a native unified understanding and generation architecture, supporting native 8K image generation, text rendering of up to 2000 characters per image, generation of up to 10 consistent-style images from a single prompt, and long-image generation such as e-commerce detail pages. It also offers 110+ preset color grading styles, 70+ preset commercial fonts, and support for seven languages.

    Why it matters: The new model unifies visual understanding and image generation in the same autoregressive framework, and its native 8K and long-text rendering capabilities are worth creators evaluating as a replacement for existing image generation workflows.

Nov 17Tue
  1. note · bochibochi_ai · 图像视频创作分享41

    Google Pics updated to Nano Banana 2.1, reducing degradation of text within images and localized edits

    On November 16, Google announced that its image generation and editing tool Google Pics has been updated to the Nano Banana 2.1 model. The new model maintains the same processing speed while improving image quality and realism, and enhancing editing consistency and text rendering; during some edits, elements such as backgrounds are less prone to degradation. The update requires no user settings and is available to paid plans such as Google AI Pro/Ultra, rolling out gradually over 1–3 days starting November 16.

  2. note · ai_freak · 图像视频创作分享58

    Seedream 5.5 update experience sharing: four improvements including realistic texture and local editing

    Seedream 5.5 update brings four improvements: more natural realistic texture, more stable local editing, clearer small text, and transparent background generation. The author tested it on Higgsfield and found that local editing can modify only the specified area without affecting other parts, and transparent backgrounds can directly generate LINE stickers and logos.

Nov 16Mon
  1. IT之家69

    ByteDance Seedance 2.5 4K video generation model API is now fully available

    ByteDance Volcano Engine announced that the Seedance 2.5 4K API is now fully available, and developers and enterprise customers can call this 4K video generation model through the interface. The model supports Draft (sample) mode, where end users can first use 480P samples to quickly validate storyboards, repeatedly refine shots, and generate the 4K final film once they are satisfied; it also features 10-bit color depth, making the image quality more suitable for the big screen.

    Why it matters: With the Seedance 2.5 4K API now open, developers can directly call the 4K generation capability, and Draft mode provides a low-cost path for storyboard refinement.

  2. Zyphra · Zonos 语音模型66

    Zyphra releases ZONOS3-Mini multilingual TTS model

    Zyphra releases compact multilingual TTS model ZONOS3-Mini, supporting 43 languages, with approximately 100 million active parameters per token, and provides a native transformers port. The model uses ZAC-1 codec as audio tokenizer, includes 27 built-in release voices, supports voice cloning from short audio, and is licensed under Apache-2.0.

    Why it matters: ZONOS3-Mini is released as a native transformers port, supporting 43 languages and voice cloning, facilitating evaluation of its suitability as a multilingual TTS baseline.

  3. 短剧自习室 · 文章75

    Tested Seedance 2.5 flash overnight; the industrialization of AI short dramas is looking more and more promising!

    Short Drama Study Room tested Seedance 2.5 flash overnight. Results show its generation speed doubled, pricing is on par with Seedance 2.0, and output quality is close to Seedance 2.5. Ten gradient tests covered micro-expressions, complex movements, 30-second long narratives, and comic-drama effects. A comparative review shows image quality close to Seedance 2.5, with half the time and lower credit consumption, making it suitable for batch production.

    Why it matters: Ten gradient tests compared Seedance 2.0/2.5/2.5 flash. The flash version takes half the time and consumes fewer credits, providing a reference for batch production selection.

Nov 15Sun
  1. MindStudio Blog · 创作分享61

    MiniMax H3 Image Review: Text, Editing, and Realism Benchmark Tests

    MiniMax H3 Image performs well in text, editing, and realism evaluations, with large text, counting, and editing identity preservation performing well; small text and complex physics are weaknesses. The model is a 33B-parameter diffusion Transformer, 6-step sampling, native approximately 16 megapixel output, supports up to 16 reference images, local ComfyUI operation requires about 48GB VRAM. In editing tests, car swaps, outfit changes, and illustration-to-photo conversions all preserved composition and identity, but storyboard physics and motion details remain insufficient.

  2. SuSu_酥酥38

    Qwen 4.0 Max Preview's pelican test results are stunning.

    Qwen 4.0 Max Preview delivers stunning results in the latest pelican test, with significant improvements in anatomical details, riding posture, scene consistency, text rendering, and lighting and materials. Market expectations were previously low, but the latest test shows the model's detail and logical abilities appear to be optimized. The long beak, throat pouch, scaly feathers, and tail feather details are precise, the "PELICANO" text on the frame is clear and undistorted, and the overall cartoon three-dimensional feel is strong.

Nov 14Sat
  1. 沃垠AI · 公众号公开订阅74

    Seedance 2.5 flash hands-on test: 13 clips covering 7 short-video scenarios

    The author tested Seedance 2.5 flash with 13 clips covering 7 short-video scenarios. Image quality is roughly on par with Seedance 2.5, generation speed is doubled, and API price is about 34% cheaper. The model's resolution is only 720P, with reference material limits of 10 images, 5 videos, and 5 audio clips. It suits daily updates, trend-following, and ad placement testing; Seedance 2.5 handles flagship quality.

    Why it matters: Testing shows that Seedance 2.5 flash delivers image quality close to Seedance 2.5 at about two-thirds the price, with double the speed, offering a cost-effective option for daily updates and ad placement testing.

  2. 新智元 · 公开公众号 RSS38

    Vita Dynamics releases Exo2EgoPolicy and Vbot Echo, cross-view generation model open to all

    Vita Dynamics releases cross-view generation model Exo2EgoPolicy, which can convert third-person videos to first-person perspective and output 21 keypoints for each hand. The model uses a 14-billion-parameter, 50-layer H3 Ref2VA DiT backbone; hand keypoint PCK@5% increased from 11.52% to 20.02%, and new-scene joint coverage is 98.92%; Vbot Echo and Echo App are now fully open.

  3. 钛媒体67

    MiniMax H3-Image Hands-On Comparison Against GPT Image 2.6 and Seedream 5.5: Faster Speed, but Unstable Execution of Complex Instructions

    MiniMax released the H3-Image model weights on November 10. The company says it can generate a 16-megapixel high-resolution image in 10 seconds, at a price about 50%-70% of mainstream tools in the same tier.

    Why it matters: The three new models show clear ranking gaps on text-to-image and image editing leaderboards. Hands-on testing shows MiniMax generates images fastest but is unstable in executing complex instructions, while OpenAI leads overall.

Nov 13Fri
  1. 电影界 · 短剧经营与内容趋势64

    Tencent WorkRally upgrade extends to short dramas, three vertical video models cut costs and boost efficiency

    Tencent's AI content creation platform WorkRally has completed an upgrade, expanding production capabilities from comic dramas to short dramas, advertisements, and creative videos, and launched three vertical video models. Among them, the video generation model Rally-Video supports multimodal references such as images, videos, and white models; the editing model Rally-Video-edit-2.0 can modify only specified local elements; and the super-resolution model UpScale-rally reduces task time by 80% compared to the previous version.

  2. 美团 LongCat · 创作模型70

    Meituan LongCat releases LongCat-Video-Avatar 1.6, an audio-driven human video generation model.

    Meituan LongCat releases the open-source audio-driven human video generation framework LongCat-Video-Avatar 1.6, which improves lip-sync, coordinated facial and body movements, and long-video consistency compared to 1.5, supports video generation over 10 minutes, and optimizes inference speed.

    Why it matters: The new version improves long-video stability and inference efficiency, and in human preference evaluation it ranks second only to OmniHuman-2.0, providing a reference for selecting digital human video generation solutions.

  3. ComfyUI Wiki · 工作流与开源创作分享63

    LongCat-Video-Avatar 1.6 released, supporting stable generation of over 10 minutes

    Meituan released the audio-driven digital human model LongCat-Video-Avatar 1.6, which supports stable generation of over 10 minutes and uses an 8-step inference path. In 33,051 manual votes, it ranked second, leading by only 1.5, about 16 Elo. ComfyUI support remains at 1.5, and 1.6 has no node path yet.

  4. Hugging Face Daily Papers24

    JVT: Just Video Transformer Makes Pixel-Space Video Diffusion Simple

    JVT (Just Video Transformer) proposes a simple large-patch DiT-style Transformer for pixel-space video diffusion, without pretrained autoencoders or U-Net-style skip connections. Its generation quality is comparable to latent diffusion baselines in both class-conditional and unconditional settings, and it supports joint generation of pixel-aligned modalities. Through camera-conditioned RGB-disparity generation, JVT improves camera following and obtains explicit scene geometry without requiring an additional reconstruction model.

Nov 12Thu
  1. Reddit StableDiffusion · 创作者与工作流分享46

    Comparison of MOSS-Image and MiniMax H3-Image on 15 identical prompts: the gap is smaller than expected

    A comparison of MOSS-Image and MiniMax H3-Image under 15 identical prompts shows that the quality gap between the two is smaller than expected. Both perform comparably on English posters, Chinese signage, portraits, and product photography, but both fail on UI mockups, and charts generate meaningless labels. The differences come more from workflow and control than from raw output quality.

  2. RenderBob · 图像视频工作流72

    Seedance 2.5 Flash released, priced at approximately 0.99 yuan/second, offering a cheaper and faster tier

    Seedance 2.5 Flash was released on November 12. A 5-second 720p no-input video example is priced at approximately 0.99 yuan/second, about 34% cheaper than Seedance 2.5 at the same specifications, but 1080p and 4K are not yet supported. The vendor claims generation speed is about twice that of Seedance 2.5; the author notes this data comes from the vendor and no independent benchmark has been seen. The API is expected to launch on Volcano Engine on November 15, and will gradually become available in applications such as Jimeng.

    Why it matters: Seedance 2.5 Flash is priced about 34% lower than Seedance 2.5, and speed is claimed to be about 2 times faster, but the author reminds that these data come from the vendor and need to be verified independently.

  3. 量子位 · 搜狐公开专栏71

    MiniMax open-sources the H3-Image image model, supporting native 4K and up to 16 reference images

    MiniMax has open-sourced the H3-Image image generation model, which supports 16MP (4096×4096) native 4K output and up to 16 reference images as input. In the author's hands-on testing, it can generate complex images such as CAD three-view drawings, mechanical exploded views, and advanced calculus exam papers, with web-based 4K generation taking about 10–20 seconds. API pricing is 50% to 70% of mainstream competitors, and 16MP requests receive an additional 50% discount during the first week.

    Why it matters: MiniMax has open-sourced the H3-Image image model, supporting native 4K and up to 16 reference images. The author's hands-on testing covers its CAD breakdowns, complex layouts, and the usability of its low-priced API.

  4. 站长之家 · AI影视与短剧产业75

    ByteDance releases Seedance 2.5 flash: generation speed doubles, 720P generation at about 0.99 yuan/second.

    ByteDance Volcano Engine releases Seedance 2.5 flash text-to-video model, with generation speed doubled compared to Seedance 2.5, and 720P generation price at about 0.99 yuan/second, down about 34% from Seedance 2.5. The model supports up to 30 seconds per segment, offers two resolutions: 480P and 720P, and the API will be available on Volcano Ark on November 15.

    Why it matters: Seedance 2.5 flash enters batch production scenarios at a 720P generation price of about 0.99 yuan/second, offering a cost-effective solution for high-frequency content production.

  5. Hugging Face Daily Papers37

    How PixVid lifts latent-space video diffusion into pixel space

    PixVid adapts a pretrained latent-space video model to generate directly frame-by-frame in pixel space, replacing the VAE decoder with a convolutional head. Fine-tuned from Wan2.1-14B, the method approaches the latent-space baseline with about 1.3M training clips and generates 2× higher resolution with the same number of tokens per frame. It maintains alignment for depth-conditioned editing under fast motion and can respond to each action in the next frame.

Nov 11Wed
  1. 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.

  2. Reddit comfyui · 创作与工作流分享68

    MiniMax-H3-Image released, ComfyUI already has native workflows.

    MiniMax releases MiniMax-H3-Image, and ComfyUI already provides two native workflows. A single checkpoint supports both text generation and image-reference editing, generating approximately 16MP per pass with six-step sampling, without the need for separate upscaling. ComfyUI needs to be updated to 0.45, and the model files include the Qwen3-VL 32B text encoder, image VAE, and Qwen3.5 9B prompt expander.

    Same story, the pick is 《MiniMax H3 Image receives day-0 native support in ComfyUI》

Nov 10Tue
  1. MiniMax · 音视频模型58

    MiniMax releases MiniMax-H3-Image-PE prompt expansion model

    MiniMax releases MiniMax-H3-Image-PE, fine-tuned from Qwen3.8-27B, providing prompt expansion and aspect ratio recommendations for MiniMax-H3-Image. It supports two workflows: text-to-image and image editing. The image editing PE is still under development; only its prompt is released this time. The model runs with FP4 weights and can be deployed via SGLang or Diffusers, requiring approximately 19 GiB of GPU memory.

  2. Pika26

    Pika H3 Flex inference lands on AMD GPU

    Many thanks to @shiyi_c98 for helping us bring Pika H3 Flex inference to AMD GPU 💜 We look forward to collaborating with more researchers in the community. If you are an AMD developer, we would love for you to contribute to this branch! Check out her work: https://github.com/Pika-Labs/Pika-H3-Flex/tree/contrib/amd-mi350x

  3. Zyphra · Zonos 语音模型65

    Zyphra releases open-source speech model ZONOS3-Mini

    Zyphra releases open-source compact speech model ZONOS3-Mini, with approximately 100M active parameters, capable of real-time operation on phones, laptops, and browsers. The model supports cloning voices from audio samples, designing characters by age/accent/timbre, and word-by-word control of emotion, speaking rate, and expression. Weights are open-sourced under Apache 2.0, and it can be served via SGLang-Omni on NVIDIA GPUs with an OpenAI-compatible interface.

    Why it matters: With 100M active parameters, it can synthesize speech in real time in the browser, providing a locally runnable open-source solution for localized dubbing and character voices, supporting voice cloning and word-by-word emotion control.

Nov 9Mon
  1. 字节 Seed 官方博客63

    Seedream 5.5 released: supports 100 rounds of stable editing and native RGBA output.

    ByteDance released the next-generation image generation model Seedream 5.5, which supports 100 rounds of multi-round editing, native RGBA transparent background output, and improved image realism and text rendering. Compared with Seedream 5.0 Pro, the usability rate across six production scenarios has improved, with film/TV and office scenarios exceeding twice the previous version; the severe text corruption rate dropped to 0.4%.

    Why it matters: Seedream 5.5 focuses on 100 rounds of stable editing and native RGBA output. Evaluations show the severe text corruption rate dropped to 0.4%, providing more reliable image assets for professional visual production.

Nov 8Sun
Nov 7Sat
  1. z-ab · 开源短剧与创作项目分享74

    OpenBMB open-sources VoxCPM2: 30 languages, 48kHz audio quality, voice design and cloning completely free

    Tsinghua OpenBMB team open-sources VoxCPM2 speech synthesis model, 2 billion parameters, supports 30 languages and 48kHz audio quality, free for commercial use under Apache-2.0 license. The model uses a diffusion autoregressive zero-tokenizer architecture, supports natural language voice design, controllable cloning and ultimate cloning, and achieves English WER 1.84% and Chinese CER 0.97% on the Seed-TTS-eval benchmark.

    Why it matters: Open-source TTS model VoxCPM2 is free for commercial use under Apache-2.0 license, supports 30 languages and 48kHz audio quality, creators can evaluate whether to replace existing paid voice-over solutions.

  2. Hugging Face Daily Papers44

    MotionGraphicsBench: Evaluating Motion Graphics Generation with Programmatic Verification

    The MotionGraphicsBench benchmark contains 100 instances and is used to evaluate large language models' ability to generate motion graphics from static SVG and motion editing prompts. On a balanced test set of 180 animations, its programmatic verification achieved a raw agreement rate of 93.3% with human judgment, higher than the GPT-6 Astra VLM baseline's 83.3%.