MarkTechPost
AI news source. Articles are auto-selected and adapted by Hamidun News editors.
Latest publications
WebBrain: open-source browser AI agent runs without the cloud or a subscription
WebBrain is a free MIT-licensed agent for Chrome and Firefox that reads pages and automates tasks, running locally via llama.cpp or Ollama.
xAI launches /goal in Grok Build: an autonomous agent plans and verifies multi-step tasks
Grok Build now has a /goal mode: the agent receives a single goal, builds its own execution plan, works through a checklist, and verifies the result at every stage until full completion.
MoonMath AI releases open-source HIP kernel for AMD MI300X that outperforms official AITER v3
Startup MoonMath AI has open-sourced an attention kernel for AMD MI300X that beats AMD's official library across all tensor shapes and rounding modes.
Cisco open-sources FAPO: automatic LLM pipeline optimization with step-by-step diagnostics
Cisco Foundation AI has open-sourced FAPO — a Claude Code-based tool that finds the weak link in a prompt chain on its own and optimizes the pipeline to a target accuracy.
TimeCopilot: How to Build a Forecasting Pipeline with Foundation Models and Anomaly Detection
TimeCopilot lets you build a complete forecasting pipeline — from statistical to foundation models — with automatic anomaly detection and an LLM agent that explains predictions.
Docling Parse: PDF parsing pipeline with character- and coordinate-level precision
The tutorial shows how to use Docling Parse for structured PDF analysis: coordinates for every word, visual overlays, and export to JSON and CSV for RAG systems.
MiniMax released MSA: sparse attention speeds up 1M-token context processing by 28x
MiniMax unveiled MSA, a two-branch sparse attention mechanism trained on a 109B-parameter MoE model that is 28x faster than standard GQA at a context length of 1M tokens.
Cohere's North Mini Code: open-weight 30B MoE model for agentic coding
Cohere has released its first open-weight model for developers — North Mini Code, built on MoE architecture with 30B parameters, a 256K context window, and the ability to run on a single H100 GPU.
Microsoft SkillOpt: automatic prompt optimization instead of manual trial and error
Microsoft SkillOpt optimizes AI prompts through an automatic six-step loop — reflection, selection, validation gating — and shows measurable accuracy gains as early as the first iterations.
Google DeepMind released DiffusionGemma — an open 26B MoE model with 4x faster generation
Google DeepMind released DiffusionGemma, an experimental 26B-parameter open model that uses text diffusion instead of autoregression — generation is up to 4x faster on GPU.
PyGraphistry: Graph analytics for threat detection in enterprise data
Researchers published a PyGraphistry Colab workflow that turns enterprise access data into an interactive graph with ML-based anomalies and three visualization modes for incident investigation.
Alibaba launched Qwen3.7-Plus: multimodal agent with video analysis and autonomous coding
Alibaba released Qwen3.7-Plus, a multimodal agent, on the Bailian platform — the model can process images and video, invoke tools, and iterate on code on its own without operator involvement.
JetBrains released Mellum2 — an open 12B MoE model for fast tasks in AI pipelines
JetBrains open-sourced Mellum2 under Apache 2.0 — a compact MoE model with 12 billion parameters, trained on 10.6 trillion tokens for integration into multi-model AI pipelines.
How to speed up transformer training with NVIDIA Apex: FusedAdam, FusedLayerNorm, and torch.amp
A step-by-step guide to building NVIDIA Apex from source and using FusedAdam, FusedLayerNorm, and mixed precision — up to 2.5× faster transformer training.
Anthropic Releases Claude Fable 5 and Mythos 5: One Model, Two Access Tiers
Anthropic opened Fable 5 to the public with full safety classifiers, while Mythos 5—with cyber restrictions removed—rolls out only through the closed Project Glasswing program.
How to build an AI agent in Google Colab without frameworks: a tutorial with memory and MCP
A step-by-step tutorial on building a lightweight AI agent in Google Colab — with tool registration, session memory, skills, and an MCP server, without external frameworks.
OpenAI unveils GPT-5.6 with three tiers — Sol, Terra, and Luna — and reasoning modes
OpenAI has announced the GPT-5.6 family with three models — Sol, Terra, and Luna — and new reasoning modes, max and ultra. Access is currently limited to selected developers.
Perplexity launches Computer for Counsel — an AI agent for legal teams
Perplexity has launched a specialized AI agent for lawyers: routing across 20+ models, integration with Microsoft 365 and Midpage, and all answers backed by verifiable source links.
NVIDIA Open-SWE-Traces: data preparation for fine-tuning coding agents
The tutorial shows how to work with the NVIDIA Open-SWE-Traces dataset — parse agent trajectories, analyze code patches, and build an SFT dataset for training AI agents.
DeepSeek released DSpark — a framework that speeds up generation on DeepSeek-V4 by 57–85%
DeepSeek open-sourced DSpark — a speculative decoding framework for DeepSeek-V4 that speeds up generation by 57–85% without loss of quality.
Meta Unveils Astryx — A React Design System with AI Agent and MCP Support
Meta has released Astryx, an open-source React and StyleX design system with an MCP server and CLI, where engineers and AI agents work with the same components via a unified API. *Meta is recognized as an extremist organ
Liquid AI released LFM2.5-230M: 213 tokens/s on Galaxy S25 and support for llama.cpp
Liquid AI has released LFM2.5-230M, an open 230M-parameter model that runs offline on a smartphone and Raspberry Pi, outperforming instruction-following rivals four times larger.
Datalab Releases lift — a 9B Open-Weight Model for Extracting JSON from PDFs
Datalab has open-sourced lift, a 9-billion-parameter vision model that converts PDFs and images into structured JSON according to a given schema, achieving 90.2% field accuracy on a 225-document benchmark.
Z.ai releases GLM-5.2: real million tokens and two levels of deep thinking
Z.ai launched GLM-5.2 with a truly working context window of 1 million tokens, High and Max modes, and support for Claude Code, Cline, and OpenClaw via Anthropic-compatible API.
FineWeb without downloading terabytes: streaming, filtering, and tokenization of web corpus for LLM
A practical guide to FineWeb from Hugging Face: how to work with a multi-terabyte web corpus for LLM training through streaming, filtering, deduplication, and tokenization — without fully downloading the data.
Zyphra Released Zamba2-VL: Visual Models with 10x Faster Response
Zyphra released Zamba2-VL — open VLMs with 1.2B, 2.7B and 7B parameters with a hybrid Mamba2 + Transformer architecture that reduces time to first token by approximately 10x.
Moonshot AI Launches Kimi Work — Desktop Agent with Swarm of 300 Sub-Agents
Kimi Work from Moonshot AI runs locally on macOS and Windows, controls the browser via WebBridge, and coordinates a swarm of up to 300 parallel sub-agents.
Claude Code 2026: Complete Breakdown of 25 Agentic Tool Features with Examples
MarkTechPost released a guide to Claude Code 2026 with a breakdown of 25 features — from CLAUDE.md and skills to MCP, hooks and Auto Mode — complete with code examples and demo.
Google releases Gemini-SQL2: Gemini 3.1 Pro scores 80% on BIRD benchmark
Google Research introduced Gemini-SQL2 based on Gemini 3.1 Pro — the model achieved 80.04% accuracy on the BIRD benchmark for text-to-SQL conversion.
How to Build an Agent Workspace on QwenPaw with Custom Skills and Streaming API
Step-by-step tutorial for deploying QwenPaw — a framework for creating AI agents with support for multiple model providers, custom skills, and streaming API in Google Colab.