From the archive
AI models are the systems behind chatbots, coding assistants and tools that work with text, images, audio and video. Follow new language model releases, open-weight models, local AI and benchmark results. Our coverage looks at capabilities, context limits, access, pricing and evaluation methods so you can compare the evidence behind a launch, not just the claims in its announcement.
Lead story
Reflection AI has introduced Beam, its first open-weight model, and is pitching reasoning efficiency rather than a clean sweep of the model rankings. The reported results put it near an earlier GLM model on one coding test, while newer Chinese models score higher on several comparisons.
Read the report & discussion ↗Watch Desk
Developer Brandon Thomas says he used Anthropic’s Claude Opus 5.5 to build seven open-source apps modelled on Adobe tools, in an ambitious project that puts AI-assisted software development and the boundaries of imitation in the same frame.
NVIDIA Watch
NVIDIA and Microsoft have announced a partnership to build hardware and software for AI agents running on Windows PCs. The plan combines NVIDIA’s RTX Spark platform with Microsoft’s agent-containment tools, bringing local AI computing and agent execution into the same Windows push.
Watch Desk
A malware campaign called PoeLLM is targeting exposed AI and development servers, using compromised machines to spread and mine cryptocurrency, according to BleepingComputer. The report says the activity has been running since at least April, putting a practical security question in front of anyone running these tools: are…
Watch Desk
Singapore’s financial regulator wants firms to have every AI use case independently reviewed before it goes live. The guidance also calls for ongoing monitoring, including of third-party AI, and says firms remain accountable for AI used in the services they provide.
Watch Desk
TypeSafe AI has emerged from two years in stealth with Jev, a model designed to give software typed decisions and confidence estimates rather than conversational answers. Its pitch is a different kind of AI tool for automation, with the company claiming striking speed and cost advantages on its…
Google DeepMind Watch
Google DeepMind’s Philipp Schmid says its managed agent environments let developers give an AI agent a persistent place to run code, install tools and continue work across requests. The practical pitch is fewer bits of infrastructure for developers to assemble themselves, and a shift from chatting with a…
AMD Watch
AMD’s ROCm 10.1 release adds faster storage-to-GPU data paths, NUMA-aware host memory and a broader set of tools for deploying and profiling AI workloads. For teams running models on AMD GPUs, the changes target the unglamorous work of getting data to the hardware and understanding where a workload…
Watch Desk
The UK AI Security Institute has released Transect, an open-source tool for exploring long AI-agent evaluation transcripts. It turns a sprawling run into an interactive timeline, helping reviewers connect an agent’s activity, tool use and delegation to the evidence behind the analysis.
Watch Desk
ASUS has announced ProArt laptops powered by NVIDIA’s RTX Spark platform, aimed at running AI models and creative workflows locally. The P16 and P14 are listed for US pre-orders from $2,599.99, with the first units due on 16 October; the announcement also outlines a compact desktop system for…
OpenAI Watch
OpenAI says it has launched GPT-6 Sol and GPT-6 Luna globally for free and paid ChatGPT users, replacing GPT-5.6 models. Its safety hub classifies both as high capability in cybersecurity and biological and chemical domains, a consequential designation alongside the company’s claimed safety improvements.
DeepSeek Watch
DeepSeek has published two new open-source repositories targeting Huawei’s Ascend accelerators, for matrix multiplication and mixture-of-experts communication. Their familiar API shapes could make experimenting beyond Nvidia’s CUDA ecosystem less of a rewrite, although the supported hardware is limited.
Stanford HAI Watch
Stanford HAI researchers found that paralegals used an AI tool 58% more, and experimented with it 70% more, when managers framed it as a way to make work more meaningful rather than simply faster. Their study suggests that training, time to explore and measures that recognise higher-value work…
Microsoft AI Watch
Microsoft is adding local AI models and sandboxed tool execution to GitHub Copilot on Windows. The change could let coding work move between a model running on a user’s PC and cloud models, while containing some agent-run commands in operating-system-level sandboxes.
Microsoft AI Watch
Microsoft says its FORGE Lab used AI-assisted research to find 140 Windows vulnerabilities between May and September 2026, including 52 addressed in September’s security release. The company’s account also describes a growing open-source vulnerability programme, where the work of validating and fixing findings may be harder to scale…
Anthropic Watch
Anthropic has launched Claude Haiku 5.5, a small model it says is designed for high-volume, cost-sensitive work such as summaries and classification. Its published API pricing starts at $0.10 per million input tokens and $0.50 per million output tokens, with higher rates for prompts over 100,000 tokens.
Microsoft AI Watch
Microsoft Research Asia has released Agent Lightning v1.0, an open-source framework for training AI agents with reinforcement learning while keeping the agent harness used in deployment in the loop. Its coding-agent example reports a 14.6 percentage-point improvement on SWE-bench Verified, with about 6,000 training samples.
Watch Desk
Liquid AI has released two open d1 decision models, including one it says can answer a question in under 50 milliseconds on each of four tested NVIDIA devices. The models return structured decisions in a single pass rather than generating text, offering developers a smaller, faster option for…
