ReasoningCerebras
GLM-4.7-REAP-218B-A32B
218B parameter MoE model with 32B active params, vLLM compatible, REAP methodology for enhanced reasoning.
Open source ↗Top AI News Weekly
Week 3 · Jan 12 – Jan 18, 2026
Speech AI dominates with Pocket-TTS, VibeVoice-7B, and PersonaPlex-7B. ShowUI-Aloha/π advance GUI agents, Google launches Universal Commerce Protocol, and GLM-Image brings hybrid image generation.
36 launches and research drops that matter for enterprise AI builders—curated, tagged, and ready for your next roadmap sync.
New drops
36
Unique sources
31
Key themes
Frontier models · Agents · Infra
frontier
Frontier Models & Research
New reasoning systems, world models, and alignment papers that move the state of the art.
Image GenZ.ai
GLM-Image
Hybrid autoregressive + diffusion image model with text-to-image, editing, style transfer, and identity-preserving generation.
Open source ↗TranslationGoogle
TranslateGemma
Open translation model supporting 55 languages, 12B outperforms Gemma 3 27B baseline with multimodal capabilities.
Open source ↗LLM ResearchStanford
Verbalized Sampling
Training-free prompting strategy to mitigate mode collapse in LLMs, achieving 2-3x diversity improvement.
Open source ↗Agent LLMNVIDIA
ToolOrchestra
Small orchestrators managing models and tools for complex reasoning on Humanity's Last Exam benchmark.
Open source ↗Small TTSekwek
Soprano-1.1-80M
Ultra-compact 80M TTS model with <1GB memory, 32kHz crystal clear audio, WebUI and OpenAI-compatible endpoint.
Open source ↗Small TTSSupertone
Supertonic-2
66M param multilingual TTS, 167× faster than real-time, optimized for on-device deployment.
Open source ↗Speech AINVIDIA
PersonaPlex-7B
Full-duplex speech model with consistent persona, based on Moshi architecture for natural conversations.
Open source ↗Speech AIMicrosoft
VibeVoice-7B
Frontier open-source TTS for podcasts and multi-speaker audio with 7.5Hz continuous speech tokenizers.
Open source ↗Speech AIKyutai Labs
Pocket-TTS
Lightweight CPU-only TTS that fits in your pocket, pip install and go with 1.6B delayed streams model.
Open source ↗Speech AISilero
Silero Models
Pre-trained enterprise-grade STT/TTS models with multi-language support via PyTorch hub.
Open source ↗Audio SRysharma3501
NovaSR
Tiny 52KB audio upsampler, 16kHz→48kHz at 3500x realtime for TTS enhancement.
Open source ↗Image GenUnsloth
Qwen-Image-2512 GGUF
Top open-source diffusion model with realistic people, rich textures, and accurate text rendering via ComfyUI.
Open source ↗immersive
Immersive Media & Simulation
Video, audio, and physics-native generation techniques shaping spatial computing.
Video GenResearch
VerseCrafter
4D-aware video world model with unified control over camera and multi-object motion via GeoAdapter.
Open source ↗Image EditAI Forever
VIBE
Visual instruction-based image editor, powerful open-source framework for text-guided editing.
Open source ↗Depth EstAIGeeksGroup
AnyDepth
Simple and efficient zero-shot monocular depth estimation with reduced parameters and computational cost.
Open source ↗3D GenMeta
ShapeR
Robust conditional 3D shape generation from casual captures using Aria glasses pipeline.
Open source ↗AnimationResearch
RigMo
Unified rig and motion learning from mesh sequences with Gaussian bones and skinning weights.
Open source ↗Medical AIResearch
HeartMula
Heart-related medical AI visualization and reconstruction research.
Open source ↗3D ReconHKUST
UniSH
Unified scene and human reconstruction in a single feed-forward pass.
Open source ↗Video GenPixVerse
PixVerse V5
AI video generation platform with latest V5 model for high-quality video creation.
Open source ↗agents
Agents & Embodied Intelligence
Embodied agents learning to act in complex virtual and hybrid worlds.
GUI AgentShowLab
ShowUI-Aloha
Human-taught GUI agent that learns workflows from screen recordings with recorder, learner, planner, and actor.
Open source ↗GUI AgentShowLab
ShowUI-π
450M flow-based VLA model for continuous GUI actions, generating smooth clicks and drags from screen observations.
Open source ↗Vision AgentGetStream
Vision-Agents
Open framework for real-time video AI with Stream's edge network for ultra-low latency.
Open source ↗Agent ResearchNVIDIA Research
SLM Agents Research
Research on small language models as the future of agentic AI for specialized applications.
Open source ↗RL EnvMeta PyTorch
OpenEnv
Interface library for RL post-training with environments including Echo, Code Sandbox, and Oumi integration.
Open source ↗EducationAdam Maj
Tiny-GPU
Minimal GPU design in Verilog to learn GPU architecture from ground up, 10k+ stars educational project.
Open source ↗tooling
Developer Tooling & Infra
Frameworks, playbooks, and OSS repos that level-up AI engineering velocity.
MCPPhil Schmid
MCP-CLI
Lightweight CLI for MCP servers with JSON output, agent-optimized for Gemini CLI and Claude Code.
Open source ↗Claude Toolthedotmack
Claude-Mem
Claude Code plugin for persistent memory across sessions, captures tool usage and injects context.
Open source ↗ProtocolGoogle
Universal Commerce Protocol
Open protocol for agentic commerce, co-developed with Shopify, Stripe, Walmart, and 20+ partners.
Open source ↗TutorialDailyDoseOfDS
MLOps Crash Course
Comprehensive MLOps + LLMOps tutorial series covering foundations to production deployment.
Open source ↗Promptinglabeldekho
Perplexity Prompts Guide
Battle-tested prompting strategies for RAG-based search engines like Perplexity AI.
Open source ↗PDF Toolpikepdf
pikepdf
Python library for reading and writing PDFs, powered by QPDF with Pythonic interface.
Open source ↗CLI Toolsupreme-gg-gg
Instagram-CLI
Terminal UI for Instagram in TypeScript/Python, the ultimate weapon against brainrot.
Open source ↗ComfyUIComfyUI Studio
ComfyUI Installers
1-click installers with UV for 100x faster installs, Torch 2.9, CUDA 13, FaceID, and IP-Adapter.
Open source ↗Resource Listad-si
Awesome Video Production
Curated list of video production tools, AI generators, teleprompters, and editing software.
Open source ↗