All AI updates

Smartsheet MCP on AWS has launched the Universal Model Context Protocol (MCP) server.

Smartsheet launches the Universal Model Context Protocol (MCP) server, enabling AI assistants to read and write manage Smartsheet data through a standardized protocol; supports Amazon Quick and Kiro connections, Quick for knowledge workers and operations teams to naturally query and update Smartsheet with natural language, Kiro allows developers to refer to project specifications and update task status without leaving the coding environment.

OpenAI introduces the complementary metric g-mean² to enhance model monitorability evaluation

OpenAI has introduced the complementary metric g-mean² to improve model monitorability evaluation. g-mean² calculates TPR and TNR through behavioral labels and monitoring predictions, but cannot solve the problems of mixed differences in different model tasks and deployment risks. The new metric includes a paired matching metric for aligned shared instances and a post-control integrated metric simulating monitoring rejection of outputs, to more comprehensively evaluate model deployment risks.

PyTorch Foundation recruits nine new members to promote the development of the open-source AI ecosystem

PyTorch Foundation announces the addition of nine new members, including Nebius, Nscale, and Shopify, strengthening its core position in the development of open-source AI and accelerating ecosystem breakthroughs. The addition of new members highlights the ecological importance of projects such as PyTorch, vLLM, and DeepSpeed, promoting the integration of AI research and commercial applications.

How does ALTK-Evolve surpass state-of-the-art models in the τ²-bench customer service task through learning and memory?

ALTK-Evolve uses GPT-4.1 to learn and apply customer service scripts. In the τ²-bench customer service agent benchmark test, it surpasses all retail entries in the retail sector, achieving a score of 67.5% Pass^4, and does not require updating model weights. The model is based on a basic ReAct agent and GPT-4.1. It attempts to analyze tasks and use GPT-4.1 to distill success and failure into short behavioral rules, then adds these guidelines to the agent's system prompts without updating weights or manual annotation.

Sakana AI Engineer Open House 2026 Event Report

Sakana AI held its second Engineer Open Day, covering five job categories, introducing the work content of each department, demonstrating applications in the finance and defense sectors, and emphasizing AI sovereignty strategy and product line expansion.

Mistral AI has launched the Voxtral 3 voice agent SDK and cutting-edge voice models.

Mistral AI has launched Voxtral 3, which includes an advanced voice model capable of real-time multi-step conversations in over 50 languages, along with a voice agent SDK. Its voice-to-text model, Voxtral Small Transcribe 3, achieved a Forced WER of 0.9% on the FLEURS 45 benchmark.

llm-d 1.0 Released: Strengthening Open Source AI Inference Production Environment

The llm-d project announces the release of version 1.0, which is a Kubernetes-native distributed inference service network that supports deploying any model's inference stack on any accelerator in any cloud. This version enhances multi-tenant quality of service, zero-downtime operations, and agent-aware services to achieve continuous performance in common production operations. Through techniques such as prefix cache-aware routing and KV cache offloading, llm-d 1.0 can capture 97% of available cache reuse, reducing the cost of large-scale AI services.

Microsoft introduces Sovereign Cloud to enhance digital sovereignty control and choice

Microsoft has launched a new phase of Sovereign Cloud, providing tiered control options through Sovereign Public Cloud, Sovereign Private Cloud, and Sovereign Partner Ecosystem to meet the digital sovereignty needs of different workloads.

AnyLanguageModel 1.0 is released: supporting local model advancements on Apple platforms for a year.

AnyLanguageModel 1.0 is released, supporting local model advancements on Apple platforms for a year. New support added for cloud models such as OpenAI Chat Completions, Anthropic Messages, and Google Gemini. Local model support includes llama.cpp, MLX, and Core ML. Tool calling and guided generation features are expanded. Cloud models have the same functionality as local models, and API keys support dynamic acquisition through closures.

Google has launched the TPU-based vLLM and SGLang inference engines.

Google has released the native PyTorch backend for vLLM and SGLang inference engines, supporting TPU and optimizing production-level services; vLLM-TorchTPU achieved a cost of 2.01 cents per million tokens for 175B parameter models in the SemiAnalysis InferenceX benchmark, leading the industry.

How does Claude design protein binders for scorpion neurotoxins?

Claude autonomously generated and screened 318,484 protein designs in 64 hours, selecting 180 for testing, and collaborated with Adaptyv Bio and Twist Bioscience to experimentally validate 211 designs, of which 17 met the binding criteria. All design data has been made public. Experts consider this the first AI design attempt for scorpion neurotoxins, which could accelerate drug discovery for neglected diseases.

Meta's TorchTitan integrates TorchTPU and optimizes TPU model training.

Meta's TorchTitan integrates TorchTPU, enabling PyTorch models to be trained natively on Google Cloud TPU with minimal code changes.

PyTorch 2.15 brings FP8 and faster LLM decoding to the MPS backend.

The Hugging Face PyTorch team introduces updates to the MPS backend in PyTorch 2.15, adding the tensor torch.float8_e4m3fn and F.scaled_mm, with FP8 tensors occupying half the space of BF16.

Anthropic uses Claude to reduce code's attack surface

How do the open standards of the Coral NPU unify edge AI and achieve portable, energy-efficient intelligence?

Coral NPU has released a complete 32-bit architecture, including new matrix execution units, supporting up to 1 TOPS performance and 10mW average power consumption; launched the Coral Compiler v1.0 based on IREE/MLIR.

Arm used Google AI Edge Model Explorer to add visual graph debugging features to custom edge compilers.

Arm leveraged the hierarchical graph visualization and plugin architecture of Google AI Edge Model Explorer to develop open-source adapters supporting formats such as TOSA, VGF, and PTE, helping developers more easily inspect edge model conversion processes.

Hugging Face launches TorchTPU: Native PyTorch backend support for TPU

Hugging Face has released TorchTPU, making TPU a native backend device for PyTorch. Transformers, Diffusers, and Kernels can now run directly on TPU. Users only need to select a model from the Hugging Face Hub and specify device_map="tpu".

Meta introduces GenRecs Tokenization to reshape ad recommendations with a stable model vocabulary.

GenRecs Tokenization enhances conversion rates by 0.35% for advertisers by representing items like ads and products with a stable semantic vocabulary, combined with Meta's Generative Ad Model (GEM) and adaptive ranking model. This technology leverages the world knowledge of pre-trained large models to enable recommendation systems to understand preferences of new items or users with no historical interaction, accelerates learning through shared concepts, achieves cross-activity transfer, and improves performance as large model scale increases, forming a feedback loop for performance improvement.

How Meta engineers improve the reliability of AI agents by streamlining routers.

Meta engineers built a lean router with 1,300 lines to assign over 20 workflows to 14 specialized AI agents, with the router itself not performing any domain-specific tasks, significantly improving the efficiency and accuracy of data pipeline agents.

Sakana AI proposes 'QD-Rec' to improve recommendation quality and diversity with an open-ended exploration method

Sakana AI proposed 'QD-Rec', recommending candidate-hypothesis pairs as units, with generative AI writing new hypotheses based on past excellent hypotheses in a cycle to accumulate high-quality and diverse recommendations. Experiments were conducted in M&A, customer targeting, and paper recommendations, achieving high average evaluations in many settings and building large sets of recommendations.

Meta's Prometheus Supercluster: The Journey to Build a Megawatt-Scale AI Infrastructure

Microsoft has launched the Project Perception system, which monitors and governs enterprise agents in real time.

Microsoft has released Project Perception, a system that uses red, blue, and green agent teams to monitor the behavior of internal agents in real-time, dynamically calculates risks, and implements deterministic control. The system utilizes a new cybersecurity stack to assign identities, container policies, and permissions to agents, and continuously monitors each action, providing real-time responses and traceability, enabling defenders to operate at the same speed as attackers. Model transparency is the next key step, requiring a deep understanding of the consistency between agent intent and task objectives.

LiteRT Tensor API release: A new solution for native model development on edge devices.

LiteRT Tensor API is released, supporting developers to write and optimize models natively in C++ on edge devices, enabling cross-platform deployment (CPU/GPU/NPU) without script conversion. Developers can fully control the computational graph structure and directly load model weights, achieving zero-copy data transfer through shared accelerator buffers, while providing building blocks for AI coding agents.

TorchNeuron: It has never been easier to run HF models on AWS Trainium.

TorchNeuron provides a new native backend for PyTorch on AWS Trainium, with native support for Transformers, Diffusers, and Kernels, making it unprecedentedly easy to run models on Trainium and share models on HF Hub.

Build a 'never silent' voice agent using ADK and Gemini 3.8.

Using ADK and Gemini 3.8, non-blocking voice interaction is implemented through streaming tools and asynchronous functions, reducing the first response time of the sales report workflow from 14.4 seconds to 1.2 seconds, supporting zero silence, progress updates, and parallel conversations.

MiniMax M3.1 Flash released: Next-generation model with improved performance and reduced cost

MiniMax M3.1 Flash is officially launched, with nearly a threefold increase in online inference speed and a 70% reduction in cost compared to the previous generation. The model performs excellently in benchmark tests for complex intelligent tasks (cowork), computer use, software development, and network security, ranking highly in the AutomationBench-AA for automated tasks, scoring 73.7% on DeepSWE v1.1, 61.7 on OSWorld V2, 29.6 on ALE, and 82% on CyberGym.

Together Post-Training: Flexible and efficient model fine-tuning service

Together AI has launched Post-Training services, supporting the construction of custom models through reinforcement learning and supervised fine-tuning without complex infrastructure. The service provides flexible training loop definition capabilities and optimizes the underlying system and infrastructure to improve model quality and training speed. It supports mainstream open-source models such as Muse Spark Open, GLM 5.3, Kimi K3, and Qwen 3.8-27B, running on dedicated GPU resources to ensure efficient and numerically consistent training.