Researchers just open-sourced tttLRM, a novel 3D reconstruction model that uses Test-Time Training to convert multiple photos into crisp 3D Gaussian Splats. Unlike traditional 3DGS, it strips away foggy artifacts to capture fine wires and tiny text, all within a compact sub-4GB model that runs on consumer GPUs.
Samsung just integrated Perplexity into Galaxy AI on the Galaxy S26, letting users summon it with "hey, Plex" right alongside Bixby and Gemini. Rather than a basic web wrapper, it gets deep OS access to native apps like Notes, Gallery, and Calendar. Why it matters: Samsung is betting on a multi-agent ecosystem where users choose specialized AI for different tasks instead of being locked into a single assistant.
Figure just put a fleet of seven Figure 03 humanoids into 24/7 autonomous operation at its California headquarters, powered by its Helix 02 neural network. The bots monitor their own battery telemetry, stepping onto 2 kW inductive charging docks and swapping tasks without human intervention. It marks a serious technical milestone for humanoid uptime, even if Figure hasn't disclosed what actual work they're doing yet.
OpenAI is quietly building a 200-person hardware team led by Apple veterans to launch consumer AI devices. Its first product is reportedly a $200 to $300 smart speaker packed with cameras for facial recognition and proactive life nudges, slated for February 2027 at the earliest. Smart glasses and a lamp are also in the pipeline.
Apple is reportedly accelerating work on three AI wearables: smart glasses codenamed N50, a camera-equipped pendant, and AI-powered AirPods. All three will tether to the iPhone and lean heavily on Siri with visual context. Production on the glasses could start as early as December ahead of a 2027 launch.
NVIDIA just dropped a guide on running the popular OpenClaw AI agent entirely for free on GeForce RTX GPUs and DGX Spark hardware. By pairing local runners like Ollama or LM Studio with open models, users can let an always-on agent manage local files and schedules without bleeding cash on cloud API tokens or handing over private data. It brings powerful autonomous desktop workflows right to consumer hardware.
Moonshine AI just released Moonshine Voice, an open-weight speech-to-text model engineered for real-time on-device use on iOS, Android, and Raspberry Pi. Its 200M parameter streaming model outperforms Whisper Large v3 in word error rate while featuring built-in voice activity detection and intent recognition.
Monarch RT just brought real-time AI video generation to consumer hardware, clocking 16 FPS on a single Nvidia RTX 5090. Its structured Monarch matrix attention runs up to 4x faster than standard FlashAttention while retaining about 95% of full-dense visual quality. Why it matters: developers no longer need multi-GPU server clusters to run live video diffusion models.
NetEase Youdao just open-sourced LobsterAI, an MIT-licensed desktop agent for Windows and macOS. It natively edits Office documents, drives an in-app browser, and takes remote jobs via chat apps like WeChat and Telegram. It brings practical, approval-gated computer workflows directly to local desktops without complex terminal setups.
Waymo just kicked off fully autonomous operations with its 6th-generation Driver on the Ojai platform. The revamped hardware relies on custom silicon to cut the camera count by more than half, slashes lidar expenses, and boosts perception in harsh winter conditions. Rides are currently limited to employees and their guests, with public access coming soon.
LavaSR just showed up as an ultralight audio enhancement tool that cleans up noisy speech almost instantly. Weighing just 50MB, it runs up to 5,000x real-time on GPUs and 60x on CPUs, making it small and fast enough to run locally on mobile phones. Why it matters: It makes real-time, zero-lag voice cleanup accessible without needing heavy cloud compute.
Soul-AILab just open-sourced SoulX-Singer, a zero-shot singing voice synthesis model. Feed it a few seconds of raw speech and a reference melody, and it makes that voice sing whatever you want without fine-tuning. The best part: at under 3GB, it runs smoothly on basic GPUs and even CPUs.
Open-source tool Just Dub It just released a lightweight LoRA adapter for the LTX-2 foundation model, letting you dub videos into languages like French and German while jointly generating matching lip movements. It preserves the speaker's facial identity and runs on consumer GPUs, with code and weights available on GitHub and Hugging Face.
Open-source developers just launched PicoClaw, an ultra-lightweight AI assistant framework designed to replace bulky setups like OpenClaw. Built from scratch in Go, it uses under 10MB of RAM and boots in less than a second on $10 single-board computers. You can hook it up to Telegram or WhatsApp to handle tasks 24/7 without burning cash on hardware.
Mistral AI just dropped Voxtral Transcribe 2, introducing two speech-to-text models for batch and live audio. The standout is Voxtral Realtime, a 4B-parameter model released under Apache 2.0 that runs on-device with latency down to sub-200ms. It brings near-offline accuracy to real-time voice agents across 13 languages without sending private audio to the cloud.
OpenBMB just rolled out MiniCPM-o 4.5, a 9B omnimodal model that sees, listens, and speaks simultaneously in real time. Thanks to an end-to-end full-duplex design, it doesn't just wait its turn—it monitors inputs at 1Hz and can proactively interrupt with reminders. It matches Gemini 2.5 Flash on key multimodal tasks while running locally on a Mac or single GPU.
Zhipu AI just open-sourced GLM-OCR, an ultra-compact 0.9B multimodal model built for heavy-duty document understanding. Despite its small size, it topped OmniDocBench V1.5 with a 94.62 score, parsing intricate tables, formulas, and handwriting into Markdown or structured JSON. It runs out of the box on vLLM and Ollama under an MIT license, making local document extraction fast and dirt-cheap.
Omnimat Zero just open-sourced a video tool that erases objects along with their tricky reflections and ground shadows. It can also isolate foreground subjects into transparent layers for background swaps, making clean video cleanup far less tedious.
ACE Studio and StepFun just open-sourced ACE-Step 1.5, a commercial-ready music foundation model built for consumer hardware. It generates full songs in under 10 seconds on an RTX 3090, supports 50+ languages, and runs with less than 4GB of VRAM. It basically puts capable music generation directly on your desktop under an MIT license.
LM Studio just launched version 0.4.0, separating its core engine into a headless daemon named llmster that runs on cloud servers and CI pipelines without a GUI. The update also brings continuous batching for parallel inference via llama.cpp and a stateful REST API supporting local MCP tools. Why it matters: Local model testing can now scale seamlessly from personal laptops to remote headless rigs.
