AI models and tools worth knowing about

[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.

open source, apache-2.0

https://github.com/thu-ml/SageAttention

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Postshot (Jawset)

Tue, 25 Aug 2026 00:00:00 +0000

Postshot, a Windows desktop application for building photorealistic 3D reconstructions from footage.

unclear, the site carries a pricing page · needs Windows 10 or later, Nvidia RTX 2060 or better

https://www.jawset.com/

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Poly Haven (free 360 images)

Tue, 25 Aug 2026 00:00:00 +0000

A free library of 360-degree environment photographs, usable for any purpose without an account.

free to use

https://polyhaven.com/hdris

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

NVIDIA Lyra

Tue, 25 Aug 2026 00:00:00 +0000

Project Lyra: Open Generative 3D World Models.

open source, apache-2.0

https://github.com/nv-tlabs/lyra

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

MoGe (Microsoft)

Tue, 25 Aug 2026 00:00:00 +0000

[CVPR'25 Oral] MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision.

source published, licence unclear

https://github.com/microsoft/MoGe

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Matrix-3D Project Page

Tue, 25 Aug 2026 00:00:00 +0000

Turns one 360-degree photograph into a 3D scene you can move around in. Project page.

open source

https://matrix-3d.github.io/

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Matrix-3D Paper

Tue, 25 Aug 2026 00:00:00 +0000

The research paper behind Matrix-3D, on generating explorable 3D worlds from a single panorama.

paper only

https://arxiv.org/abs/2508.08086

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

✓ Matrix-3D (Skywork AI)

Tue, 25 Aug 2026 00:00:00 +0000

Generate large-scale explorable 3D scenes with high-quality panorama videos from a single image or text prompt.

open source, mit

✓ We have tried this one.

https://github.com/SkyworkAI/Matrix-3D

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Marble (World Labs)

Tue, 25 Aug 2026 00:00:00 +0000

A commercial tool for creating and sharing 3D worlds in the browser.

hosted service · needs none, runs in the cloud

https://marble.worldlabs.ai/

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

LightX2V

Tue, 25 Aug 2026 00:00:00 +0000

Lightweight Image Video Action Generation Inference Framework.

open source, apache-2.0

https://github.com/ModelTC/LightX2V

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

LichtFeld Studio

Tue, 25 Aug 2026 00:00:00 +0000

Train, inspect, edit, automate, and export 3D Gaussian Splatting scenes from a single native application.

open source, gpl-3.0

https://github.com/MrNeRF/LichtFeld-Studio

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Krea 2 Technical Report

Tue, 25 Aug 2026 00:00:00 +0000

Krea's image generation models, released with downloadable weights in a standard and a fast version.

open weights

https://www.krea.ai/blog/krea-2-technical-report

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

HY-World 2.0 (Tencent Hunyuan)

Tue, 25 Aug 2026 00:00:00 +0000

HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worlds.

source published, licence unclear

https://github.com/Tencent-Hunyuan/HY-World-2.0

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Brush

Tue, 25 Aug 2026 00:00:00 +0000

3D Reconstruction for all.

open source, apache-2.0

https://github.com/ArthurBrussee/brush

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Apple SHARP Project Page

Tue, 25 Aug 2026 00:00:00 +0000

Apple's method for turning one photograph into a viewable 3D scene in under a second, with paper and code published.

open source

https://apple.github.io/ml-sharp/

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Apple SHARP Paper

Tue, 25 Aug 2026 00:00:00 +0000

Apple's paper on reconstructing a viewable 3D scene from a single photograph in under a second.

paper only

https://arxiv.org/abs/2512.10685

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

Apple SHARP

Tue, 25 Aug 2026 00:00:00 +0000

Sharp Monocular View Synthesis in Less Than a Second.

source published, licence unclear

https://github.com/apple/ml-sharp

Named in We Open Sourced World Generation (Mickmumpitz, 2026-08-25).

SenseNova U1.5

Sun, 23 Aug 2026 00:00:00 +0000

An 8-billion-parameter reasoning model from SenseTime, distributed through ModelScope.

open weights

https://modelscope.ai/models/SenseNova/SenseNova-U1.5-8B-MoT

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Qwen Video Edit

Sun, 23 Aug 2026 00:00:00 +0000

Edits video by written instruction, including long clips where each segment gets a different instruction.

research project

https://yunpeng1998.github.io/Qwen-Video-Edit-Page

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Ornith 1.5

Sun, 23 Aug 2026 00:00:00 +0000

A model family built to improve its own answers in a loop rather than in a single pass.

open source

https://ornith.ai/ornith_1_5.html

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Happy Shrimp

Sun, 23 Aug 2026 00:00:00 +0000

A music service that writes a complete song, lyrics and vocals included, from a one-line prompt.

hosted service · needs none, runs in the cloud

https://www.happyshrimp.ai/

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

GeoWeaver

Sun, 23 Aug 2026 00:00:00 +0000

Builds a consistent 3D reconstruction from long video sequences without drifting off course.

research project

https://kosmoresearch.github.io/GeoWeaver/

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Gen 1.5

Sun, 23 Aug 2026 00:00:00 +0000

A robot model that picks up a new task from a few seconds of demonstration, or a few minutes of data.

hosted service

https://generalistai.com/blog/gen-1.5

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Evoke

Sun, 23 Aug 2026 00:00:00 +0000

Named in the video, but the project page it linked to no longer exists.

unavailable

https://evoke-world.github.io/

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Deepseek vision

Sun, 23 Aug 2026 00:00:00 +0000

DeepSeek's image understanding: send a picture with your question and it reads charts, screenshots and text.

paid service · needs none, runs in the cloud

https://api-docs.deepseek.com/guides/vision

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Comfy MCP

Sun, 23 Aug 2026 00:00:00 +0000

Lets a coding assistant drive a ComfyUI install on your own machine, using your GPU, your models and your custom nodes.

open source · needs your own GPU

https://blog.comfy.org/p/open-sourcing-comfy-mcp-on-local

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Bernini v2

Sun, 23 Aug 2026 00:00:00 +0000

Model weights (bernini, image-text-to-video) published on Hugging Face as ByteDance/Bernini-Diffusers-v2.

open, apache-2.0

https://huggingface.co/ByteDance/Bernini-Diffusers-v2

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

✓ Avo

Sun, 23 Aug 2026 00:00:00 +0000

NVIDIA's argument that the scaffolding around a model, not the model alone, decides how well an agent performs on long tasks.

paper only

✓ We have tried this one.

https://developer.nvidia.com/blog/nvidia-avo-reaches-100-on-arc-agi-3-demonstrating-a-frontier-level-general-purpose-architecture-for-long-horizon-autonomous-agents/

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

Audio8 TTS 0.1B

Sun, 23 Aug 2026 00:00:00 +0000

Speech synthesiser (arktts, feature-extraction) published on Hugging Face as Audio8/Audio8-TTS-Preview-0.1b.

open weights under the publisher's own licence — read it

https://huggingface.co/Audio8/Audio8-TTS-Preview-0.1b

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

4DAnyone

Sun, 23 Aug 2026 00:00:00 +0000

Rebuilds a moving 3D person from one ordinary handheld video, with no camera calibration.

research project

https://4danyone.github.io/

Named in New AI waifus, new Deepseek, realtime worlds, Happy Shrimp, tiny TTS: AI NEWS (AI Search, 2026-08-23).

WorldClaw

Sun, 16 Aug 2026 00:00:00 +0000

Turns a single open-ended prompt into a 3D world you can walk through and edit.

research project

https://tencent-hunyuan.github.io/Hunyuan3D-WorldClaw/

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Sign to text

Sun, 16 Aug 2026 00:00:00 +0000

DeepMind's sign-language-to-text model, now driving features in shipping products.

hosted service

https://deepmind.google/blog/putting-sign-language-ai-into-users-hands/?utm_source=chatgpt.com

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Scope

Sun, 16 Aug 2026 00:00:00 +0000

Gives an existing video model precise camera control by telling it which direction each part of the frame is being viewed from.

research project

https://visual-ai.github.io/scope

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Qwen 3.8 max

Sun, 16 Aug 2026 00:00:00 +0000

Language model (qwen3_5_moe_text, text-generation) published on Hugging Face as Qwen/Qwen3.8-2.4T-A95B.

open weights under the publisher's own licence — read it

https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

✓ Qwen 3.8 27B

Sun, 16 Aug 2026 00:00:00 +0000

Vision-language model (qwen3_5, image-text-to-text) published on Hugging Face as Qwen/Qwen3.8-27B.

open, apache-2.0

✓ We have tried this one.

https://huggingface.co/Qwen/Qwen3.8-27B

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Nemotron Lightning

Sun, 16 Aug 2026 00:00:00 +0000

A small open NVIDIA model plus a routing library, meant to run the same workload across a laptop, a workstation or a data centre.

open weights

https://blogs.nvidia.com/blog/nemotron-lightning-switchyard-rtx-dgx

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Muse Glimmer

Sun, 16 Aug 2026 00:00:00 +0000

Meta's 30-billion-parameter model for agents that run continuously on your own machine, tuned for tool use and recovering from mistakes.

open weights, apache-2.0 · needs a single GPU

https://developer.meta.com/ai/models/muse-glimmer/

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

MiDasheng

Sun, 16 Aug 2026 00:00:00 +0000

Generates full audio scenes of variable length by treating sound the way a language model treats words.

research project

https://xingws.github.io/midashenglm-gen-demo/

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

MatrAIx

Sun, 16 Aug 2026 00:00:00 +0000

Testing infrastructure that simulates a range of users against your product before you release it to real ones.

open source

https://matraix.ai/

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Magi 2

Sun, 16 Aug 2026 00:00:00 +0000

Model weights (text-to-video, image-to-video) published on Hugging Face as sand-ai/MAGI-2-preview.

open, apache-2.0

https://huggingface.co/sand-ai/MAGI-2-preview

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

LTX 2.5

Sun, 16 Aug 2026 00:00:00 +0000

Generates multi-shot video in one pass, edits real footage, and exports at cinema quality. Weights are downloadable.

open weights · needs runs on your own hardware

https://ltx.io/model/ltx-2-5

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

JoyAI video edit

Sun, 16 Aug 2026 00:00:00 +0000

[Official Repo] JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion.

open source, apache-2.0

https://github.com/jd-opensource/JoyAI-Video-Edit

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Index TTS 2.5

Sun, 16 Aug 2026 00:00:00 +0000

Speech synthesiser (indextts, text-to-speech) published on Hugging Face as IndexTeam/IndexTTS-2.5.

open weights under the publisher's own licence — read it

https://huggingface.co/IndexTeam/IndexTTS-2.5

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Grok 4.6

Sun, 16 Aug 2026 00:00:00 +0000

xAI's assistant model, this version aimed at agents that run for a long time without supervision.

paid service · needs none, runs in the cloud

https://x.ai/news/grok-4-6

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

GPT Ultrafast

Sun, 16 Aug 2026 00:00:00 +0000

An OpenAI service tier that runs their model far faster on specialised chips, quoted at up to 750 words per second.

paid service · needs none, runs in the cloud

https://openai.com/index/previewing-ultrafast/

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

GLM 5.3

Sun, 16 Aug 2026 00:00:00 +0000

The announcement page for Zhipu's GLM 5.3 language model. The page returned no readable text to us; the downloadable weights are listed separately in this archive.

unclear

https://z.ai/blog/glm-5.3

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Gemini 3.7 Flash

Sun, 16 Aug 2026 00:00:00 +0000

Google's mid-tier Gemini model, positioned for everyday coding and agent work.

paid service · needs none, runs in the cloud

https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash/

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Dyna 2

Sun, 16 Aug 2026 00:00:00 +0000

A robot control model trained on over a million hours of ordinary human video, presented with evidence that it keeps improving with scale.

hosted service

https://www.dyna.co/dyna-2

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Deepseek V4 0813

Sun, 16 Aug 2026 00:00:00 +0000

Language model (deepseek_v4, text-generation) published on Hugging Face as deepseek-ai/DeepSeek-V4-Pro-0813.

open, mit

https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Deepseek harness

Sun, 16 Aug 2026 00:00:00 +0000

DeepSeek Harness: Everything is a Plugin.

open source, mit

https://github.com/deepseek-ai/deepseek-harness

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Cactus Needle

Sun, 16 Aug 2026 00:00:00 +0000

A tiny 45-million-parameter model for calling tools and pulling structured data out of text, small enough to ship inside an app.

open weights · needs runs in about 28 MB of memory

https://cactuscompute.com/needle

Named in New Deepseek, GLM 5.3, Grok 4.6, LTX 2.5, Qwen 3.8, Gemini 3.7: AI NEWS (AI Search, 2026-08-16).

Muscriptor

Sat, 15 Aug 2026 00:00:00 +0000

The link points at a publisher's Hugging Face account rather than a specific release, and the page could not be read without signing in.

unclear

https://huggingface.co/MuScriptor

Named in The BEST local AI music generator is here! (AI Search, 2026-08-15).

Minimax Music for Comfy

Sat, 15 Aug 2026 00:00:00 +0000

Official instructions for running MiniMax Music 3 inside ComfyUI to produce complete songs up to five minutes long.

free to use

https://docs.comfy.org/tutorials/audio/minimax/minimax-music-3

Named in The BEST local AI music generator is here! (AI Search, 2026-08-15).

Smaller models

Wed, 12 Aug 2026 00:00:00 +0000

Model weights published on Hugging Face as Kijai/MiniMax-H3-experimental.

unclear

https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

Realism lora

Wed, 12 Aug 2026 00:00:00 +0000

Model weights (minimax-h3, lora) published on Hugging Face as fal/MiniMax-H3-Realism-People-LoRA.

open weights under the publisher's own licence — read it

https://huggingface.co/fal/MiniMax-H3-Realism-People-LoRA

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3 Turbo Lora

Wed, 12 Aug 2026 00:00:00 +0000

Video generator (minimax-h3, text-to-video) published on Hugging Face as larryvrh/MiniMax-H3-Turbo-Lora.

open, apache-2.0

https://huggingface.co/larryvrh/MiniMax-H3-Turbo-Lora

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3 Turbo

Wed, 12 Aug 2026 00:00:00 +0000

Video generator driven by a still image (minimax-h3, comfyui) published on Hugging Face as joyfox/MiniMax-H3-Turbo.

unclear

https://huggingface.co/joyfox/MiniMax-H3-Turbo

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

Minimax h3 Turbo

Wed, 12 Aug 2026 00:00:00 +0000

Video generator driven by a still image (t2v, i2v) published on Hugging Face as lightx2v/Minimax-h3-Turbo.

open, apache-2.0

https://huggingface.co/lightx2v/Minimax-h3-Turbo

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3 prompt guide (reference)

Wed, 12 Aug 2026 00:00:00 +0000

Model weights (minimax-h3, text-to-video) published on Hugging Face as MiniMaxAI/MiniMax-H3.

open weights under the publisher's own licence — read it

https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/docs/VIDEO_PROMPT_WRITING_GUIDE_ref_en.md

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3 prompt guide (basic)

Wed, 12 Aug 2026 00:00:00 +0000

Model weights (minimax-h3, text-to-video) published on Hugging Face as MiniMaxAI/MiniMax-H3.

open weights under the publisher's own licence — read it

https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/docs/VIDEO_PROMPT_WRITING_GUIDE_base_en.md

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3 GGUF

Wed, 12 Aug 2026 00:00:00 +0000

Video generator driven by a still image (gguf, comfyui) published on Hugging Face as Abiray/MiniMax-H3-GGUF.

open weights under the publisher's own licence — read it

https://huggingface.co/Abiray/MiniMax-H3-GGUF

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3 comfy

Wed, 12 Aug 2026 00:00:00 +0000

Model weights published on Hugging Face as Kijai/MiniMax-H3_comfy.

unclear

https://huggingface.co/Kijai/MiniMax-H3_comfy/

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

MiniMax H3

Wed, 12 Aug 2026 00:00:00 +0000

Model weights (minimax-h3, text-to-video) published on Hugging Face as MiniMaxAI/MiniMax-H3.

open weights under the publisher's own licence — read it

https://huggingface.co/MiniMaxAI/MiniMax-H3

Named in This new AI Render Engine is amazing! (Mickmumpitz, 2026-08-12).

Live preview

Wed, 12 Aug 2026 00:00:00 +0000

Model weights published on Hugging Face as Kijai/MiniMax-H3-TAE.

open, apache-2.0

https://huggingface.co/Kijai/MiniMax-H3-TAE

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

Image workflow

Wed, 12 Aug 2026 00:00:00 +0000

Model weights published on Hugging Face as reverentelusarca/minimax-h3-comfyui-workflows.

open weights under the publisher's own licence — read it

https://huggingface.co/reverentelusarca/minimax-h3-comfyui-workflows

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

GGUFs

Wed, 12 Aug 2026 00:00:00 +0000

Model weights (gguf, text-to-video) published on Hugging Face as unsloth/MiniMax-H3-GGUF.

open weights under the publisher's own licence — read it

https://huggingface.co/unsloth/MiniMax-H3-GGUF

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

Easy workflow

Wed, 12 Aug 2026 00:00:00 +0000

The easiest way to use MiniMax H3. One compact workflow for T2V, I2V, first/last-frame, and reference video generation, with a unified multi-media input, powerful @ references, and inline dialogue blocks. Less wiring. More creative control.

open source, mit

https://github.com/nkxx188/ComfyUI-MiniMaxH3-Easy

Named in The BEST local AI video generator just got BETTER! (AI Search, 2026-08-12).

Workflows and files

Wed, 05 Aug 2026 00:00:00 +0000

Official ComfyUI instructions for MiniMax H3, covering text, image and reference-driven video with stereo sound.

free to use

https://docs.comfy.org/tutorials/video/minimax/minimax-h3

Named in The BEST local AI video generator is here! Minimax H3 tutorial (AI Search, 2026-08-05).

Spectrum

Wed, 05 Aug 2026 00:00:00 +0000

Training-free Spectrum acceleration for ComfyUI’s native MiniMax H3 audio-video model. Uses Chebyshev ridge feature forecasting to skip selected H3 transformer evaluations, with adaptive scheduling, sampler-aware support for Euler, ER-SDE, RES, SEEDS and SA-Solver, CPU/VRAM history storage, and fail-closed native fallbacks.

open source, gpl-3.0

https://github.com/xmarre/ComfyUI-Spectrum-MiniMax-H3

Named in The BEST local AI video generator is here! Minimax H3 tutorial (AI Search, 2026-08-05).

Sage attention wheels

Wed, 05 Aug 2026 00:00:00 +0000

Fork of SageAttention for Windows wheels and easy installation.

open source, apache-2.0

https://github.com/woct0rdho/SageAttention/releases

Named in The BEST local AI video generator is here! Minimax H3 tutorial (AI Search, 2026-08-05).

More on license

Wed, 05 Aug 2026 00:00:00 +0000

Model weights (minimax-h3, text-to-video) published on Hugging Face as MiniMaxAI/MiniMax-H3.

open weights under the publisher's own licence — read it

https://huggingface.co/MiniMaxAI/MiniMax-H3/discussions/12

Named in The BEST local AI video generator is here! Minimax H3 tutorial (AI Search, 2026-08-05).

✓ Minimax H3 models

Wed, 05 Aug 2026 00:00:00 +0000

Model weights (diffusion-single-file, comfyui) published on Hugging Face as Comfy-Org/MiniMax-H3.

open weights under the publisher's own licence — read it

✓ We have tried this one.

https://huggingface.co/Comfy-Org/MiniMax-H3

Named in The BEST local AI video generator is here! Minimax H3 tutorial (AI Search, 2026-08-05).

KJ nodes

Wed, 05 Aug 2026 00:00:00 +0000

Various custom nodes for ComfyUI.

open source, gpl-3.0

https://github.com/kijai/ComfyUI-KJNodes

Named in The BEST local AI video generator is here! Minimax H3 tutorial (AI Search, 2026-08-05).

Wonder

Sun, 02 Aug 2026 00:00:00 +0000

A world model that generates an interactive scene from a starting image or video.

research project

https://wonder-world-model.github.io/

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ Seedance 2.5

Sun, 02 Aug 2026 00:00:00 +0000

ByteDance's video model that generates picture and sound together, built for 30-second stories with reference images and editing.

paid service · needs none, runs in the cloud

✓ We have tried this one.

https://seed.bytedance.com/en/seedance2_5

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Redesign

Sun, 02 Aug 2026 00:00:00 +0000

Official repository for "ReDesign" (ECCV 2026).

source published, licence unclear

https://github.com/jintae-00/ReDesign

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Prism

Sun, 02 Aug 2026 00:00:00 +0000

A compact way of representing which of a robot's own signals actually matter for the task it is doing.

research project

https://lsh3163.github.io/prism/

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Phi Zero

Sun, 02 Aug 2026 00:00:00 +0000

A world model that learns how physical scenes change by watching ordinary video.

research project

https://phi-zero.github.io/

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ Minimax H3

Sun, 02 Aug 2026 00:00:00 +0000

An all-in-one generation model that reads text, images, video and audio together, and produces video with stereo sound at up to 2K and 15 seconds.

open weights

✓ We have tried this one.

https://www.minimax.io/blog/minimax-h3

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Kimi K3

Sun, 02 Aug 2026 00:00:00 +0000

Vision-language model (kimi_k3, feature-extraction) published on Hugging Face as moonshotai/Kimi-K3.

open weights under the publisher's own licence — read it

https://huggingface.co/moonshotai/Kimi-K3

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ Instella

Sun, 02 Aug 2026 00:00:00 +0000

AMD's fully open 16-billion-parameter language model, of which only 2.8 billion are used per word, trained on AMD's own accelerators.

open weights

✓ We have tried this one.

https://rocm.blogs.amd.com/artificial-intelligence/instella-moe/README.html

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Inkling Small

Sun, 02 Aug 2026 00:00:00 +0000

Vision-language model (inkling_mm_model, image-text-to-text) published on Hugging Face as thinkingmachines/Inkling-Small.

open, apache-2.0

https://huggingface.co/thinkingmachines/Inkling-Small

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Ideogram obj remover

Sun, 02 Aug 2026 00:00:00 +0000

An online tool for removing unwanted objects from a photograph. The page blocks automated readers, so nothing beyond the name was confirmed.

hosted service · needs none, runs in the cloud

https://ideogram.ai/apps/object-remover

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ ID V2V

Sun, 02 Aug 2026 00:00:00 +0000

Restyles a video into a different look while keeping the person recognisably themselves. From Netflix and Eyeline Labs.

research project

✓ We have tried this one.

https://eyeline-labs.github.io/ID-V2V/

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Gemini voice typing

Sun, 02 Aug 2026 00:00:00 +0000

Dictation in the Gemini app for macOS: speak, and it produces cleaned-up text, edits and summaries.

hosted service · needs none, runs in the cloud

https://blog.google/innovation-and-ai/products/gemini-app/speak-naturally-gemini-app-mac-os/

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

Gemini Robotics 2

Sun, 02 Aug 2026 00:00:00 +0000

DeepMind's robot model extending control from the hands to the whole body.

research project

https://deepmind.google/blog/gemini-robotics-2-brings-whole-body-intelligence-to-robots/

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ Deepseek V4 Flash 0731

Sun, 02 Aug 2026 00:00:00 +0000

Language model (deepseek_v4, text-generation) published on Hugging Face as deepseek-ai/DeepSeek-V4-Flash-0731.

open, mit

✓ We have tried this one.

https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ Crisper Whisper

Sun, 02 Aug 2026 00:00:00 +0000

Speech recognition that transcribes exactly what was said, in many languages, with the timing of each individual word.

open weights, non-commercial only

✓ We have tried this one.

https://nyra-labs.com/crisperwhisper

Named in New Deepseek, Seedance 2.5, Minimax H3, Gemini Robotics, AMD models: AI NEWS (AI Search, 2026-08-02).

✓ ShotPlan

Sun, 26 Jul 2026 00:00:00 +0000

Generates multi-shot video with clean cuts and camera moves at the exact frames you ask for.

open source

✓ We have tried this one.

https://pensioner-11.github.io/ShotPlan/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Sana Video 2

Sun, 26 Jul 2026 00:00:00 +0000

NVIDIA's efficient 720p video generator, built to cut the cost of the attention step.

research project

https://nvlabs.github.io/Sana/Video2/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Qwen 3.8

Sun, 26 Jul 2026 00:00:00 +0000

The paid subscription page for Alibaba's Qwen cloud service.

paid service · needs none, runs in the cloud

https://www.qwencloud.com/pricing/token-plan

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Opus 5

Sun, 26 Jul 2026 00:00:00 +0000

Anthropic's top assistant tier, aimed at agents that run for long stretches and at professional coding work.

paid service · needs none, runs in the cloud

https://www.anthropic.com/news/claude-opus-5

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

OpenDreamer

Sun, 26 Jul 2026 00:00:00 +0000

An openly released world model with a playable demo, published together with an account of how it was trained.

open source

https://next-state.github.io/open-dreamer/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

OpenAI hack

Sun, 26 Jul 2026 00:00:00 +0000

OpenAI and Hugging Face's joint account of a security incident during model evaluation. News, not a tool.

news article

https://openai.com/index/hugging-face-model-evaluation-security-incident/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

New Gemini models

Sun, 26 Jul 2026 00:00:00 +0000

Three new Gemini tiers, including a lighter one and a security-focused one.

paid service · needs none, runs in the cloud

https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-6-flash-3-5-flash-lite-3-5-flash-cyber/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

✓ Nanbeige

Sun, 26 Jul 2026 00:00:00 +0000

A 3-billion-parameter reasoning model, small enough for a single consumer graphics card.

open weights · needs one consumer GPU

✓ We have tried this one.

https://modelscope.ai/models/nanbeige/Nanbeige4.2-3B

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Mage Flow

Sun, 26 Jul 2026 00:00:00 +0000

Named in the video, but the Hugging Face page could not be read without signing in, so it may be restricted or withdrawn.

unclear

https://huggingface.co/microsoft/Mage-Flow

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Laguna S2.1

Sun, 26 Jul 2026 00:00:00 +0000

poolside's coding model, this version aimed at work that takes many steps to finish.

hosted service · needs none, runs in the cloud

https://poolside.ai/blog/introducing-laguna-s-2-1

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Homie

Sun, 26 Jul 2026 00:00:00 +0000

Puts a specific person and object into generated video while keeping both recognisable.

research project

https://yiyangcai.github.io/homie-page.github.io/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

GPT live in desktop

Sun, 26 Jul 2026 00:00:00 +0000

Voice conversation in the ChatGPT desktop app.

hosted service · needs none, runs in the cloud

https://learn.chatgpt.com/docs/features/voice

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Google quantum breakthrough

Sun, 26 Jul 2026 00:00:00 +0000

Google's work on using reinforcement learning to improve quantum error correction. News, not a tool.

news article

https://research.google/blog/towards-a-quantum-computer-that-learns-from-its-errors/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

GLM with vision

Sun, 26 Jul 2026 00:00:00 +0000

Vision-language model (sglang, glm5v) published on Hugging Face as baseten/GLM-5.2-Vision-NVFP4.

open, mit

https://huggingface.co/baseten/GLM-5.2-Vision-NVFP4

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

Flux 3

Sun, 26 Jul 2026 00:00:00 +0000

Black Forest Labs' single model covering images, video, audio and predicting actions.

unclear

https://bfl.ai/models/flux-3

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).

ChatGPT Health

Sun, 26 Jul 2026 00:00:00 +0000

Lets eligible US users connect medical records and Apple Health to ChatGPT for personalised health summaries.

hosted service · needs none, runs in the cloud

https://openai.com/index/health-in-chatgpt/

Named in Claude Opus 5, GPT 6 hack, Flux 3, new Gemini, quantum breakthrough, new Qwen: AI NEWS (AI Search, 2026-07-26).