Classification and update method
Only official primary sources are collected. OpenAI, Anthropic, and Google official updates are used as the base reference. Each update unit separates what changed, what it enables, and practical impact, so readers can compare updates consistently.
Classification and update method
Category definitions
Dynamically listed from categories.js
text / Text Generation Updates around writing, editing, and contextual text handling.
image / Image Generation Updates related to image generation, editing, and reference-based visual workflows.
video / Video Generation Updates involving video generation, editing, and time-based visual control.
voice / Voice Updates covering speech input, voice interaction, and text-to-speech output.
search / Search and Research Updates that improve web search quality, evidence use, and research workflows.
agent / Agent and Automation Updates that change how tasks are decomposed, executed, and coordinated.
api / API Updates to API-available models, invocation patterns, and integration workflows.
multimodal / Multimodal Updates involving non-text inputs and cross-modal processing.
infra / Infrastructure and Operations Updates tied to model delivery infrastructure, runtime behavior, and operational stability.
tooling / Tooling Updates for development workflows, implementation support, and adjacent tooling.
Service definitions
Dynamically listed from services.js
chatgpt / ChatGPT (OpenAI) A core service with broad coverage across research, generation, and execution support, making workflow-wide updates easier to track.
claude / Claude (Anthropic) A core service strong in long-context understanding and context retention, making writing and continuity-focused improvements easier to spot.
gemini / Gemini (Google) A core service strong in API usage and multimodal workflows, making implementation-oriented updates around the Google ecosystem easier to follow.
Update unit fields
Fixed 5-field structure
titlesummarycapabilityimpactsource