The daily radar on models, frameworks, and hardware to run AI locally. LLMs, LangChain, Chroma, mini-PCs, and everything you need for a distributed "in-house" brain.
AI-generated articles from feeds, with space for human editorial layer above the raw content.
OpenAI is facing a potential loss of internal resources due to the departure of some researchers. The news raises questions about the stability and future direction of the company, a leader in the artificial intelligence sector.
According to a Deloitte AI Institute report, companies are scaling the adoption of agentic and physical AI systems, achieving productivity gains. However, governance gaps remain, and there are difficulties in transforming pilot projects into stable systems. Interest in data sovereignty and local infrastructure is growing.
Atlassian has assured investors it can add AI to its services without blowing out its costs or shrinking margins. CEO feels under-appreciated amid year-long value slump.
A user reported a significant throughput increase, up to 26 tokens/second, using the Qwen3-Coder-Next-Q4_K_S model with llama.cpp on an RTX 5090. The optimization was achieved by offloading MoE expert tensors to the CPU and quantizing the KV cache.
Memory manufacturer PSMC reports narrowing losses, driven by rising DRAM prices and increasing demand for artificial intelligence solutions. This positive trend reflects an improving semiconductor market.
Taiwan's drone exports are surging, with the economics ministry confident in reaching the NT$20 billion target. This increase reflects the growing global demand for drones in both civilian and military applications, and Taiwan's ability to compete in the international market.
Optics manufacturer Largan reported an 11% increase in yearly revenue, despite a seasonal slowdown. The company, specializing in smartphone components, continues to benefit from demand in the sector, while still being affected by typical market fluctuations.
Cloud service providers (CSPs) are exploring custom silicio solutions to diversify their hardware options and reduce dependence on traditional vendors like Nvidia. This trend could lead to new architectures optimized for specific workloads.
OpenAI is facing increasing competition in the artificial intelligence market, while profitability continues to be an elusive goal. The article analyzes the challenges the company faces.
Google is significantly increasing its investments in AI infrastructure, particularly in TPU ASICs. This move intensifies competition among suppliers and signals a strong push towards custom hardware solutions for artificial intelligence workloads.
Foxconn reports a 35% revenue increase in January, driven by strong demand for AI servers. This reflects the growing importance of specialized hardware for AI workloads, both in the cloud and on-premise.
Taiwanese manufacturer Wistron reported an exceptionally positive January, driven by strong demand for servers dedicated to artificial intelligence. This highlights the growing market interest in specialized hardware solutions for AI workloads.
Chrome's latest revision of its browser extension architecture, known as Manifest v3 (MV3), was widely expected to make content blocking and privacy extensions less effective than its predecessor, Manifest v2 (MV2). However, this has not been the case.
A pull request introduces tensor parallelism in Llama.cpp, paving the way for faster and more efficient inference on large language models. The community welcomes this development, which could significantly improve performance on distributed hardware.
According to DIGITIMES, artificial intelligence and advanced applications (AP) are boosting shipments of load boards. January revenues show growth, indicating strong demand in the sector.
Taiwanese LED equipment maker FitTech has won an international arbitration against China's Sanan Optoelectronics, recovering NT$1.49 billion (approximately $46 million). The dispute concerned alleged contract violations. The decision highlights the importance of legal protection in international business, especially in the technology sector.
Cerebras Systems has announced a funding round that nearly triples its valuation in just six months. The company focuses on developing specialized hardware for artificial intelligence workloads, particularly for training large models.
South Korea is making significant investments in artificial intelligence, supported by a hardware infrastructure powered by over 260,000 Nvidia GPUs. This strategic move aims to position the country as a leader in the AI sector, with a focus on advanced applications and research.
According to Digitimes, Google's recent advancements in integrating artificial intelligence into its search engine demonstrate how AI is enhancing, not replacing, existing search functionalities. The company is achieving significant efficiency gains, suggesting a future where AI and traditional search coexist and reinforce each other.
The LocalLLaMA community is questioning the future of Gemma 4, wondering if Google is still investing in the development of the language model. Despite progress in the sector, the fate of Gemma 4 remains uncertain.
Reddit identifies AI-powered search as a significant growth opportunity for its business. The company aims to improve user experience and further monetize the platform through new search functionalities.
Amazon Web Services (AWS) recorded its best quarter in 13 quarters in Q4 2025. Strong demand for artificial intelligence services significantly contributed to this result, driving adoption of Amazon's cloud platform.
SoproTTS v1.5 is a 135M parameter TTS (text-to-speech) model offering zero-shot voice cloning. Trained for approximately $100 on a single GPU, the model achieves around 20x real-time speed on a base MacBook M3 CPU. The new v1.5 version offers reduced latency and improved stability.
OpenAI has announced GPT-5.3-Codex, a new version of its advanced coding model, accessible via command line, IDE extension, web interface, and a new macOS desktop app. This model outperforms previous versions in benchmarks like SWE-Bench Pro and Terminal-Bench 2.0, expanding its applications to deployment management, debugging, and test result handling.
The US Department of Justice redacted the face of the Mona Lisa in a 2009 email, part of the files related to Jeffrey Epstein. Simultaneously, sensitive data of victims were released online, raising criticism about the department's actions.
The GNU Nettle cryptographic library has a major new update that introduces support for SLH-DSA, the post-quantum signature scheme selected by NIST for the FIPS 205 standard.
OpenAI has announced Frontier, a platform designed to support enterprises in implementing software agents based on advanced models. The initiative aims to facilitate the adoption of artificial intelligence solutions in the enterprise context.
In 2015, SpaceX's ambitions to create a low-earth orbit internet-providing satellite constellation were seen as a step towards an all-encompassing company. Today, with over 9,000 satellites, SpaceX dominates the sector. The article analyzes how the ability to launch satellites at low cost gives SpaceX significant power in the telecommunications and space launch industries.
OpenAI has announced an update to its agentic coding model Codex, designed to accelerate development capabilities. The news arrives shortly after a similar announcement from Anthropic, signaling growing competition in the sector.
An autonomous lab combining OpenAI’s GPT-5 with Ginkgo Bioworks’ cloud automation cut cell-free protein synthesis costs by 40% through closed-loop experimentation. This automated approach promises to accelerate biological research and reduce development costs for new drugs and materials.
Elon Musk's plan to create orbital data center clusters dedicated to artificial intelligence seems to be taking shape. The initiative could open new frontiers for data processing in space, but also raises technical and logistical questions.
Anthropic, the maker of Claude, appears to be taking a jab at OpenAI with an ad campaign alluding to the latter's plans. AI companies are looking for new ways to spend resources, other than model training. One strategy is to buy high-profile ad space to compete with rivals.
LightOnOCR-2 and GLM-OCR, two new models for optical character recognition (OCR), have been released. A user reported superior performance compared to solutions available in late 2025, with GLM-OCR offering speed and reliable structured output.
Intel's Xe graphics driver for Linux, starting with kernel 7.0, will re-enable D3cold support for Battlemage GPUs. This feature was disabled due to instability issues in power state transitions. The change will not apply to all systems, excluding specific NUCs.
OpenAI has launched Frontier, a new platform designed for enterprises to build and deploy agents while treating them like human employees.
Introducing GPT-5.3-Codex, a Codex-native agent designed to tackle complex real-world technical tasks. It combines frontier coding performance with general reasoning capabilities to support long-horizon projects.
GPT-5.3-Codex has been unveiled, an advanced model for code generation that combines the performance of GPT-5.2-Codex with superior reasoning and professional knowledge capabilities. The model positions itself as one of the most advanced of its kind.
Meta has developed a PyTorch-based inference system for recommendations, crucial for translating advanced research into production services. The article describes the workflow, from the definition of the trained model to inference transformations, optimizations, and requirements for a high-performance inference server, focusing on the efficient use of GPUs and C++ runtime.
OpenAI introduces Trusted Access for Cyber, a trust-based framework that expands access to frontier cyber capabilities while strengthening safeguards against misuse. The initiative aims to balance innovation with responsibility in the cybersecurity field.
Anthropic has released version 4.6 of Opus, its flagship language model. This release aims to broaden its appeal to new use cases, particularly those involving AI agent teams.
DeepBrainz has released DeepBrainz-R1, a family of small language models (4B, 2B, 0.6B) focused on reasoning for agentic workflows. Optimized for multi-step reasoning and stability in tool-calling, these Apache 2.0 models aim to provide predictable behavior in local and cost-sensitive setups.
Meta is testing a standalone application for 'Vibes', its AI-generated short-form video platform. Launched last September, Vibes allows users to create and share AI videos and access a dedicated feed.
Google introduces a new framework, called NAI (Natively Adaptive Interfaces), that leverages artificial intelligence to make technology more adaptive and inclusive. The goal is to improve the user experience for everyone, regardless of their abilities or specific needs.
Nearly 60% of SAP migration projects are delayed and over budget, according to ISG research. Organizations often underestimate complexity, allow scope expansion, and fail to understand internal constraints. ECC support ends in 2027.
Debian's continuous integration (CI) infrastructure has restricted public access to its data due to excessive scraping by bots used to train large language models (LLMs). The load generated by these scrapers has impacted web server resources.
HP and Dell are reportedly evaluating DRAM from CXMT, while Acer and Asus are considering Chinese suppliers for memory chips. This move could have a significant impact on the component supply chain and PC production costs.
Microsoft says "reliability is the priority" for AI in Visual Studio. The reassurance may raise eyebrows among developers already living with Copilot's quirks.
Trillion Labs and KAIST AI introduced gWorld, an open-weight visual world model for mobile GUIs. gWorld, available in 8B and 32B versions, generates executable web code instead of pixels, surpassing larger models like Llama 4 in accuracy. This approach offers better visual fidelity and text precision compared to pixel-based or text-only models.
A Reddit user benchmarked the Strix Halo's iGPU, testing various software configurations with 13 LLM models and 15 different llama.cpp builds. The aim was to evaluate the impact of ROCm, Vulkan, and various compilation options on inference performance.
Unofficial pre-built ik_llama.cpp builds are now available for macOS, Ubuntu, and Windows. These builds simplify project adoption by removing the need for manual compilation. The creator still encourages compiling from the original source code when possible.
Google Cloud has developed an AI tool to support the U.S. Ski and Snowboard athletes. The goal is to improve their performance through data analysis and technique optimization.
Tenstorrent announced a reduction in the number of Tensor cores on its Blackhole p150 cards, from 140 to 120, via a firmware update. The company anticipates a 1-2% performance drop for existing users. New cards will ship with 120 Tensor cores.
A new online tool allows users to check if their LinkedIn contacts are mentioned in the recently unsealed Epstein files. The tool, called EpsteIn, analyzes public documents and generates a report with the findings. Accuracy is not guaranteed, but it can provide insights.
Fundamental has built a new foundation model to extract insights from enterprise structured data. The company raised $255 million in a Series A funding round to enhance its analytics platform.
ElevenLabs CEO argued at Web Summit Qatar that voice is the next interface for AI, as OpenAI, Google, and Apple push conversational systems into wearables, new hardware, and everyday interactions.
The UK government, in collaboration with Microsoft, announces a framework to evaluate deepfake detection technologies, responding to the exponential growth of AI-generated content. However, industry experts express doubts about the actual effectiveness of this initiative in stopping the proliferation of digital forgeries.
Western Digital announces the development of 3.5-inch HDDs (Hard Disk Drives) based on HAMR (Heat-Assisted Magnetic Recording) technology with a capacity reaching 140 TB, thanks to the use of 14 platters. This technology promises to significantly increase the storage density of hard drives.
Nvidia is reportedly developing DLSS 4.5, an advanced version of its upscaling technology that could eliminate the need for denoisers in ray tracing. This is thanks to a Transformer model that reconstructs ray-traced reflections more accurately.
A new report indicates that AI-powered bots already account for a meaningful share of web traffic. An increasingly sophisticated arms race is unfolding, as bots deploy clever tactics to bypass website defenses meant to keep them out.
Microsoft has made OneDrive agents generally available. Users can now query multiple documents simultaneously through Copilot, instead of just one at a time. This new feature expands Copilot's capabilities in analyzing data spread across different files.
AI-Radar is an independent observatory covering AI models, local LLMs, on-premise deployments, hardware, and emerging trends. We provide daily analysis and editorial coverage for developers, engineers, and organizations exploring local AI solutions.
