Home
Tools
Models
App
Prompts
News
Topics
Submit for inclusion
English
English 日本語 한국어 Português español Deutsch Русский Français 繁體中文 简体中文
option
Home Close
Home
Tools
Category Ranking Generate Cover Generate Image Skill
Models
Large Language Model Multimodal Model
App
Prompts
Prompts Image Prompts
News
News Flash News
Topics
Submit for inclusion
Submit App Submit Tool
Create an account Sign In
Change Language English Change Language
Settings
English EN 日本語 JA 한국어 KO Português PT español ES Deutsch DE Русский RU Français FR 繁體中文 ZH-TW 简体中文 ZH-CN
Home
Flash News
Content
ArthurYoung
ArthurYoung
July 2, 2026

AReaL 2.0, an open-source reinforcement learning infrastructure, launches to enable agents to continuously learn in real-world business scenarios. It records agent interactions, transforms them into training data, and supports safe online RL. Developed by Ant Group, Tsinghua, and HKUST, the project is now open-sourced on GitHub and linked to the PyTorch Foundation.

AReaL 2.0, an open-source reinforcement learning infrastructure, launches to enable agents to continuously learn in real-world business scenarios. It records agent interactions, transforms them into training data, and supports safe online RL. Developed by Ant Group, Tsinghua, and HKUST, the project is now open-sourced on GitHub and linked to the PyTorch Foundation.
0
0
Share Click to copy the link
Comments (0)
0/300
Submit
Author GaryWilson GaryWilson September 20, 2026
WeChat AI open-sourced WeKnora for actionable knowledge graphs. Zhipu AI fixed ZCode data privacy issues and pledged to open-source code. OpenAI launched a free ChatGPT Word plugin, joining Copilot and Claude in office AI. StepZen released Step5Preview, a cost-effective open-source model with 1/8th the cost of Claude Opus5. SpaceXAI updated Grok Voice Transcribe 2.0, halving error rates. Bilibili ranked GPT-6 top in its AI Arena. Qwen released Qwen3.8-LiveTranslate with 2.3s delay. DeepSeek offered half-price API billing during holidays and released the cost-effective V4.1 Flash model.
0
0
/live/7477
Author KeithHarris KeithHarris September 20, 2026
Over ten major AI models are set for launch around the Mid-Autumn and National Day holidays. OpenAI plans to release GPT-6 Sol, Terra, and Luna, while Anthropic launches Opus 5.2 and Fable 5.2. Google may debut Gemini 4.0 Pro and Flash. Domestically, Jieque Star released Step 5 Preview, with Kimi K3.1, Alibaba's Qwen4, and Xiaomi's Mimo V2.6 pending. Zhipu and MiniMax are rumored to upgrade to trillion-parameter models. DeepSeek's anticipated V4.1 Pro, potentially doubling parameters to 3 trillion, remains the most watched release.
0
0
Author FredBrown FredBrown September 20, 2026
WeChat AI team open-sourced WeKnora 0.8.0, a knowledge management framework designed to enable execution rather than just retrieval. It features anydoc parsing for structured data extraction and GraphRAG for reliable relationship mapping. The key innovation is Skill Sandbox Runtime, which encapsulates knowledge into executable skills within secure environments like Docker, allowing models to run verifiable tasks. Combined with long-term memory and a modular MIT-licensed architecture, WeKnora supports end-to-end workflows from parsing to execution for both individual and enterprise use cases.
0
0
Author RichardJackson RichardJackson September 20, 2026
Step launches Step5Preview, a 600B sparse MoE model optimized for real-world Agentic tasks. Activating only 27B parameters, it supports 1M token context and multimodal inputs. Scoring top-three globally, it costs one-eighth of Claude Opus5 while matching performance in coding, long-horizon reasoning, and finance. Key benchmarks include surpassing Opus5 in GPU kernel optimization and demonstrating robust multi-step agent workflows. Full open-source weights release October 15, positioning Step5Preview as a high-efficiency flagship for daily agent use.
0
0
/live/7474
Author KeithNelson KeithNelson September 20, 2026
Thunder launches AI Master M7000, a 120B parameter local AI workstation priced at 19,900 RMB. Featuring AMD Ryzen AI Max+395 and 64GB RAM, it utilizes Phison's aiDaptiv+ to treat SSD as GPU memory extension. This enables efficient local inference of large models without professional GPUs, targeting creation, development, and confidential work in a 2.45kg device.
0
0
/live/7473
OR