Every week, I organize 400 to 700 links into roughly 60 categories as part of my ongoing effort to learn about AI. This is my personal notebook, which I enjoy sharing with friends… a hobby and a labor of love, rather than a commercial publication or product.
If you arrived here through a search or shared link, this page collects the links I found for Multimodal for the week ending August 14, 2026.
As part of my learning process, I like to automate the category covers. It gives me a chance to learn Python and APIs.
This week’s cover prompt was written using Claude Opus 4.7, and the image was generated using Nano Banana 2.1.
Category Cover Image Prompt: A single photorealistic glass prism resting on a still pale surface, one thin warm gold beam entering and exiting as a soft resolved spectrum, surrounded by vast empty dawn-white and deep indigo mist, minimalist fine-art magazine composition with the bold title 'Multimodality' set large and clean in the negative space, soft diffused early light, quiet symmetry, no other text or logos.
This Week in Multimodal News
Summary by Claude Sonnet 5.5 based on this week’s links:
- Robots learning from video: Dyna Robotics introduced Dyna-2, a world-action model pre-trained on more than 1 million hours of egocentric human video that predicts next frames and actions together. The company says it found scaling laws across four orders of magnitude of human data. Separately, Stanford and NVIDIA's PointWorld-1B predicts how a whole scene moves from RGB-D input and robot actions.
- Open video and music models: LTX-2.5 launched as an open video model, and a post says it is already in Diffusers, generating video and 48 kHz sound together. MiniMax Music 3 also arrived with open weights, turning a prompt and lyrics into full songs. On the closed side, Google highlighted Gemini Omni Flash for creating and editing video from text, image, video or audio references.
- Small vision-language models: Cohere released North Micro Vision under Apache 2.0, pitching it as its smallest vision-language model, aimed at document understanding. Cohere says it beats Gemma 4 E2B and Ministral 3 3B on a range of visual benchmarks. Meta's Muse Glimmer is a 30B multimodal reasoning model, reportedly with a Gemma-like design, according to Sebastian Raschka.
This summary was generated by Claude Sonnet 5.5 to help you explore the links below. Rest assured, I select, organize, and check the links by hand in Google Sheets, and write the introduction and personal commentary in The Main Newsletters myself each week as a labor of love.
This week's links related to Multimodal
Databricks Grows >80% YoY, Surpasses $7B Revenue Run-Rate, Scales Lakebase, Genie, and Unity AI Gateway – Databricks
https://www.databricks.com/company/newsroom/press-releases/databricks-grows-80-yoy-surpasses-7b-revenue-run-rate-scales
What if a robot could simulate the physical world from a single image. [📍Bookmark Paper & GitHub for later] PointWorld-1B from Stanford and NVIDIA is a large 3D world model that predicts how an entire scene will move, given RGB-D input and robot actions. The key idea is”
https://x.com/IlirAliu_/status/2086151375084265842
WorldTrace: Addressable Memory for Video World Models
https://research.nvidia.com/labs/sil/projects/WorldTrace/
Today we are introducing Dyna-2, a world-action model pre-trained on one million hours of human video. At this scale, for the first time, we discovered several new scaling laws: • world-action models exhibit scaling law on human data across four orders of magnitude, from 1000″
https://x.com/DynaRobotics/status/2086856327150858298
world models are brutal”
https://x.com/dreamingtulpa/status/2086546779852378356
What a LiDAR sensor sees when cyclists are racing through Iowa.”
https://x.com/bilawalsidhu/status/2085788609290834241
We recently introduced Gemini Omni Flash, our first model in the new Omni family. With Omni, you can easily create and edit high-quality videos from text, image, video or audio references. We recently gave developers access to it, and since then, we’ve seen builders all over the”
https://x.com/Google/status/2086814383582118356
1 billion Gemini app users every month 1 billion downloads of the Gemma family of models Onwards!”
https://x.com/OfficialLoganK/status/2087262466703286652
As thousands of cyclists finished the 391.4-mile RAGBRAI trek across Iowa, Ouster BlueCity tracked every move through a City of Dubuque intersection. Our proprietary AI detection model was trained on 4 million labeled objects across 800 sites, covering diverse traffic patterns,”
https://x.com/ousterlidar/status/2085770236406485296
DeepSeek V4 Flash + LFM2.5-VL-3B as the vision auxiliary in Hermes Agent is a deadly combo. DeepSeek handles the agentic work. LFM2.5-VL-3B gives it the eyes. I told it to use only its local vision model to browse. It read Hacker News, picked the top story, opened the eclipse”
https://x.com/noctus91/status/2087559912687862240
Must-read papers of the week ▪️ EnvACE ▪️ The Optimizer Is the Agent: Reasoning-Driven Search across Prompts, Programs, and ML Workflows ▪️ SFT Conflicts, RL Coexists ▪️ Addressable Memory for Video World Models ▪️ The Personalization Mirage: How LLMs Fabricate User Profiles,”
https://x.com/TheTuringPost/status/2087141521493787031
Introducing LTX-2.5: The Open World Model for Video, Real-Time, & Physical AI | LTX
https://ltx.io/newsroom/introducing-ltx-2-5
Dyna-2: A 1-Million-Hour Scaling Law for World-Action Models … DYNA
https://www.dyna.co/dyna-2
How well do egocentric human videos scale? Apparently, really well. Dyna Robotics has introduced Dyna-2, a world-action model (WAM) pre-trained on more than 1 million hours of real egocentric human video. It learns by jointly predicting the next video frames and the actions, and”
https://x.com/TheHumanoidHub/status/2086875827296436394
MiniMax Music 3 is just out on @huggingface 🎶🎵 SoTA song generation, now open weights 🔊, a 8B LLM + 2.7B DiT that turn prompt + lyrics into full songs fits smol consumer GPUs with 🧨 diffusers or comfy model weights + app on Spaces, come play ▶️
https://x.com/multimodalart/status/2087933660490056163
I’m leaving Meta to start a new company. Building the TBD Lab alongside Mark @finkd and Alex @alexandr_wang has been deeply inspiring and fulfilling. I’m proud of what our multimodal team accomplished across Muse Spark, Voice Mode, Muse Image, and Muse Video, and even prouder of”
https://x.com/jiahuiyu/status/2087936732939616299?s=20
LTX-2.5 from @Lightricks is now in 🤗 Diffusers. A few cool details: > makes video + 48 kHz sound together > can pick the clip length from your prompt > uses a 2-pass mode for better quality > can render in tiles to save memory > re-compresses input images to better match”
https://x.com/RisingSayak/status/2087457946770850274
Exciting update: MiniMax-H3 by @MiniMax_AI just landed #1 in the Video Edit Arena within open models and overall! With 1390 pts, this release is +32 pts over the next two best models, Dreamina Seedance 2.0 and Gemini Omni Flash, and +83 pts than HappyHorse 1.0 at #4. Congrats”
https://x.com/arena/status/2087930695469646276
The model punches above its weight, outperforming Gemma 4 E2B and Ministral 3 3B across a broad range of visual understanding benchmarks and delivers particularly strong results on document understanding and visual Q&A. See the full results:
https://x.com/cohere/status/2087571579517489581
We are in an insane run of open-weight drops. Every modality, open source is winning. This is what an open source AI summer ☀️ looks like: 🧠 LLMs & Reasoning → DeepSeek-V4-Flash-0731 (my king 👑): 304B MoE refresh, Terminal-Bench 2.1 jumps 61.8→82.7 over the preview,”
https://x.com/victormustar/status/2087551400377037062
I really like this new benchmark. Has the flavor of ARC-AGI3 but it’s pure text so you don’t have to worry about the vision capability”
https://x.com/tri_dao/status/2087677140410290302
this week in open source: – Muse-Glimmer-30B – LTX 2.5 – Nemotron 3.5-Lightning – Ling3.0-tiny – Qwen3.8-2.4T – LFM2.5-VL-3B – Wan2.2-Animate-2 – Cohere North-Micro-Vision and there’s more coming 🤭”
https://x.com/multimodalart/status/2087576052457513234
Today, we’re adding another member to our model family. Meet North Micro Vision. Our smallest vision-language model yet, ideal for sophisticated document understanding. Available open-source under an Apache 2.0 license. Get the weights on @huggingface.”
https://x.com/cohere/status/2087571573947392419
extremely exciting to see meta getting back into open weight models with a 30B dense first, and soon muse spark 1.2 they used knowledge distillation from muse spark, architecture wise it’s similar to gemma 4 (which is llama 3 + swa (again!) + vision encoder), with scale free QK”
https://x.com/eliebakouch/status/2086769240271405477
Whoa, Meta released a new open-weight LLM yesterday, something that hasn’t happened since the good old Llama days. Their Meta Muse Glimmer model is a 30B multimodal reasoning model with a Gemma-like architecture design. (“Glimmer” is probably a wordplay on “Spark,” the more”
https://x.com/rasbt/status/2087180773497421926
SL2T is our breakthrough sign language-to-text model powering new features for Deaf and hard of hearing users on @Android. Starting with American Sign Language-to-English on Pixel 11, people can sign directly into Gboard and Live Transcribe instead of typing.”
https://x.com/GoogleDeepMind/status/2087541213284946191
OCR 4.1 – Mistral AI | Mistral Docs
https://docs.mistral.ai/models/ocr-4-1
open-weights version of Qwen3.8-Max is text-only. no vision input support. fuck.”
https://x.com/skalskip92/status/2087578544801010075
Leave a Reply