AN IN-DEPTH REPOSITORY
Every model, paper, media release, agent and tool from the month—in one source-linked archive.
COMPLETE MONTHLY ARCHIVE
Frontier coding models accelerated while video, music, voice and embodied systems moved toward faster, more controllable production use.
THE COMPLETE MONTHLY INDEX
Every linked paper, product release, research project and article has been checked directly. Cards use the publication's own description, imagery and published date wherever the source supplied them.
The AI Search helps discover the wider field, while every available card opens the original publication. Dates come only from publication metadata or official source APIs—never from a video. A checked date records our review, not the release. Older undated entries show their archive month.
PUB.SAKANA.AIInstant LLM Updates with Doc-to-LoRA and Text-to-LoRA
Recent LLM agents have shown impressive capabilities on complex computer use and long-horizon tasks. Yet, they still struggle with long-term memory and adaptation --two of the most important cognitive capabilities that still limit LLMs today. Without…
GUOXU1233.GITHUB.IODreamID-Omni
Unified Framework for Controllable Human-Centric Audio-Video Generation
ECHOESOVERTIME.GITHUB.IOEchoes Over Time: Unlocking Length Generalization in Video-to-Audio Generation Models
Christian Simon 1 , Masato Ishii 2 , Wei-yao Wang 1 , Koichi Saito 2 , Akio Hayakawa 2 , Dongseok Shim 1 , Zhi Zong 1 , Shuyang Cui 1 , Takashi Shibuya 2 , Shusuke Takahashi 1 , Yuki Mitsufuji 1,2 1 Sony Group Corporation 2 Sony AI
CODEYSUN.GITHUB.IOGenerated Reality: Human-centric World Simulation using Interactive Video Generation with Hand and Camera Control
Generated Reality: Human-centric World Simulation using Interactive Video Generation with Hand and Camera Control
GITHUB.COMGitHub - ysharma3501/LavaSR: 🌋LavaSR: Fast Speech restoration and enhancement
🌋LavaSR: Fast Speech restoration and enhancement. Contribute to ysharma3501/LavaSR development by creating an account on GitHub.
RESEARCH.NVIDIA.COM
QUIVER.AIQuiverAI – Building the Future of Vector Design
QuiverAI helps teams generate, edit, and animate editable SVG assets for product, brand, and marketing workflows.
HUGGINGFACE.COQwen3.5 - a Qwen Collection
Qwen35 is published on Hugging Face. Open the model or project card for the author’s documentation, files, license and usage details.
SOLARIS-WM.GITHUB.IOSolaris: Multiplayer video world model
Multiplayer video world model with scalable multi-agent data engine
LIANGBINGZHAO.GITHUB.IOPhysicEdit
From Statics to Dynamics: Physics-Aware Image Editing with Latent Transition Priors
CWCHENWANG.GITHUB.IOtttLRM
tttLRM: Test-Time Training for Long Context and Autoregressive 3D Reconstruction
XK-HUANG.GITHUB.IOVecGlypher | Project Page | CVPR 2026
VecGlypher: Unified Vector Glyph Generation with Language Models. Official project page summary with method, data, results, and citation.
VIDEO-REASON.COMA Very Big Video Reasoning Suite
An unprecedentedly large-scale suite for video reasoning, comprising VBVR-Dataset, VBVR-Bench, and VBVR-DataFactory, enabling systematic scaling studies and emergent generalization in video reasoning models.
TUE-MPS.ORGVidEoMT: Your ViT is Secretly Also a Video Segmentation Model
VidEoMT reveals Vision Transformers can perform video segmentation efficiently without task-specific components .
CORRER-ZHOU.GITHUB.IOHiFi-Inpaint: Towards High-Fidelity Reference-Based Inpainting for Detail-Preserving Human-Product Images
HiFi-Inpaint is a high-fidelity reference-based inpainting framework for generating detail-preserving human-product images. Accepted at CVPR 2026.
ZUNWANG1.GITHUB.IOAnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
ZEYUET.GITHUB.IOAudioX: A Unified Framework for Anything-to-Audio Generation
AudioX: A Unified Framework for Anything-to-Audio Generation
AIGEEKSGROUP.GITHUB.IOCode2Worlds: Empowering Coding LLMs for 4D World Generation
A curated gallery of Code2Worlds renders showcasing the geometry, lighting, and imagination that the model unlocks in every scene.
GITHUB.COMGitHub - KittenML/KittenTTS: State-of-the-art TTS model under 25MB 😻
State-of-the-art TTS model under 25MB 😻 . Contribute to KittenML/KittenTTS development by creating an account on GitHub.
UNICORNANROCINU.GITHUB.IOLUVE : Latent-Cascaded Ultra-High-Resolution Video Generation with Dual Frequency Experts
LUVE enables ultra-high-resolution video generation via latent-cascaded modeling with dual-frequency experts.
INFINI-AI-LAB.GITHUB.IOMonarchRT: Efficient Attention for Real-Time Video Generation
GEMINI.GOOGLELyria – Gemini AI-musikk- og sanggenerator
Lag egendefinerte lydspor for alle anledninger. Lyria 3 transformerer tekst eller bilder til lydspor med vokal og sangtekster.
HUGGINGFACE.COQwen/Qwen3.5-397B-A17B · Hugging Face
Qwen3.5 397B A17B is published on Hugging Face. Open the model or project card for the author’s documentation, files, license and usage details.