News & Updates
July 31, 2026
Midjourney V8.2 is officially launched: the "aesthetic obsession" of a 60-person laboratory
Midjourney announced the official launch of V8.2; as a 60-person research laboratory funded by the community, it continues to iterate around the five major themes of imagination, coordination, reflection, beauty and human prosperity. V8.2 continues the path of "quality and aesthetics" as its core competitiveness.
Source: Midjourney Official
Keling AI releases Kling 3.0: VIDEO 3.0 and Omni reshape the "narrative logic of light and sound"
Kuaishou Keling AI releases the KlingAI 3.0 series, with the official slogan "All in One, One for All"; VIDEO 3.0 and Omni natively support deep multi-modal command analysis and cross-task integration, achieving the dual binding of visual identity and voice intonation.
Source: Keling AI official
June 9, 2026
Midjourney Product Observation: Capability Inventory of AI Image and Design Direction
Midjourney is an AI image generation tool for the creative industry. This article organizes its core parameters, version evolution, usage methods, regional restrictions and subscription prices to facilitate quick evaluation by individual creators and teams.
Source: Midjourney official documentation
How to use Microsoft Copilot: Microsoft’s official AI assistant
Microsoft Copilot is Microsoft's consumer-grade AI assistant, based on GPT-4o and integrated with Bing real-time search, free to use.
Source: Microsoft Copilot official documentation
Mastra Capability Inventory: Selection Reference for AI Agent Teams
This article dismantles Mastra from parameters, functions, costs, versions and implementation boundaries to help the team form executable judgments during the pilot and expansion stages.
Source: Mastra official documentation
Meta AI Product Observation: Capability Inventory in the Direction of AI Chat Assistant
Meta AI is Meta's free AI assistant, based on the Llama model and built into Facebook, Instagram, WhatsApp and Messenger, supporting AI conversations, image generation and real-time search.
Source: Meta AI official documentation
Re-understanding Manus: the core capabilities of AI agent tools
Manus is a general-purpose autonomous AI agent for end users. This article sorts out its capability stack, latest 1.6 version, pricing, typical scenarios and suitable groups.
Source: Manus official documentation
Make Product Observation: Capability Inventory in the Direction of AI Office and Efficiency
Make is a visual automation platform that connects thousands of applications to orchestrate workflows. This article summarizes its functions, pricing and application scenarios.
Source: Make official documentation
Luma Ray 3.2 released: returning "directorial control" of video generation to creators
Luma released the Ray 3.2 model and API on June 9, focusing on "directly directing videos like a director" - precisely controlling each frame, keeping the subject and style consistent across shots, and outputting movie-level films. It is located at the core of the model layer of Luma's "Model + Agent + API" three-layer architecture.
Source: Luma official news
Understand Microsoft 365 Copilot in one article: Deeply integrated with the full set of Microsoft 365 applications
Microsoft 365 Copilot deeply embeds GPT-4 into the full set of M365 applications such as Word, Excel, PowerPoint, Outlook, and Teams.
Source: Microsoft 365 Copilot official documentation
What is Luma Dream Machine: Flagship AI video from Luma Labs
Luma Dream Machine is the flagship AI video model launched by Luma Labs. The Ray2 series is close to the first echelon in terms of physical realism and lens language.
Source: Luma Dream Machine official documentation
How to use Lovable: directly generate full-stack applications through natural language
Lovable allows users to quickly generate and iterate full-stack applications through natural language. This article summarizes its product capabilities, typical processes, pricing methods, and differences with tools such as Cursor and v0.
Source: Lovable official documentation
Leonardo AI function dismantling and applicable scenario analysis
Leonardo AI is an AI image generation platform produced in Australia. It supports dozens of exclusive fine-tuned models, real-time canvas, Motion video and 3D texture generation. It is aimed at creative professionals and game developers. The free version is 150 tokens per day.
Source: Leonardo AI official documentation
What is Llama: Meta open source large language model series
Llama is an open source large language model series released by Meta. Llama 3.1 provides three sizes of 8B/70B/405B, 128K context, multi-language support, and open weight for local deployment and commercial use (free for 700M MAU or less).
Source: Llama official documentation
How to use Langfuse: open source AI engineering platform
Langfuse provides a closed loop of AI engineering from prototype to production, supporting Tracing, Prompts, Evaluations, Datasets and self-hosted deployment, which is suitable for teams that require observable and evaluation governance.
Source: Langfuse official documentation
How to use LangChain: 138K+ Star’s Agent Engineering Platform
LangChain is an open source LLM application framework with 138K+ stars on GitHub. It provides chain calls, graph agent orchestration (LangGraph) and observability platform (LangSmith).
Source: LangChain official documentation
Understand Langflow in one article: Focus on improving the efficiency of high-frequency tasks
Langflow's public functions, version context, cost structure, application scenarios and applicable boundaries are organized to facilitate quick evaluation of implementation value.
Source: Langflow official documentation
Krea function dismantling and applicable scenario analysis
Krea provides real-time AI canvas and image and video generation capabilities, emphasizing creative feedback speed and visual controllability. This article summarizes its core functions, pricing structure, applicable scenarios and competitive product positioning.
Source: Krea official documentation
How to use Kling AI: the flagship AI video generation model launched by Kuaishou
Kling AI is the flagship AI video generation model launched by Kuaishou. KlingAI 3.0 ranks first in the Chinese world in terms of action coherence, camera movements, and graphic videos. This article summarizes its core functions, version evolution and pricing plan.
Source: Kling AI official documentation
One article to understand Kimi: The Dark Side of the Moon’s flagship AI assistant
Kimi is Dark Side of the Moon’s flagship AI assistant, specializing in long text processing, reasoning, and coding tasks. This article summarizes Kimi’s product capabilities, version evolution, pricing path and typical application scenarios.
Source: Kimi official documentation