[ REF — Source register & editorial policy ]
Evidence first. Footnotes included.
A publication should make its evidence easy to inspect. This register connects current model facts, tool capabilities and practical guidance to the original documents.
Source-checked Dated editorial snapshot · official sources · no invented resultsSources & corrections
80
primary documents
Official model cards, release notes, protocol specifications and developer documentation.
35
field guides
Source-linked explanations with practical procedures, checklists and clearly bounded claims.
29
tool profiles
Documentation reviews with use cases and limitations, not invented hands-on ratings.
Primary source register
Anthropic
- Introducing Claude Opus 5.5
- Claude models overview
- Claude Platform release notes
- Claude Opus 5.5 system card
- Claude Fable 5.1
- Introducing Claude Sonnet 5 — pricing correction included
- Claude Code — commands and workflows
- Claude Code — permission and configuration reference
- Claude Code — hook lifecycle and limitations
OpenAI
- OpenAI API changelog
- GPT-6 Astra: a new generation of intelligence
- GPT-6 Astra API model and pricing
- GPT-6 Sol API model and pricing
- GPT-6 Luna API model and pricing
- Codex — agent approvals and security
- Codex — local configuration and search modes
- OpenAI — current prompt-caching behavior
- OpenAI — model migration and cache changes
- Introducing Gemini 3.8 Flash and Flash Cyber
- Gemini 3.8 Flash model specifications
- Gemini 3.8 Flash — introductory pricing and migration
- Gemini API release notes
- Gemini 3.8 Live and Live Extended Thinking
- Gemini 3.8 Live model specifications
- Building with Gemini Audio — per-minute pricing
- Gemini 3.8 text-to-speech says hello
- Gemini Omni in Google Vids
- Gemini Notebook — new study tools
xAI / SpaceXAI
DeepSeek
Qwen
Mistral AI
Cursor
GitHub
Model Context Protocol
Ollama
LM Studio
Open WebUI
n8n
Dify
Flowise
LangChain
LlamaIndex
Qdrant
pgvector
Langfuse
Arize AI
promptfoo
LiteLLM
Perplexity
ElevenLabs
Runway
Midjourney
Vercel
Lovable
Replit
Aider
Correction & expansion log / September 2026
- Current models and news: replaced stale prototype releases with the reviewed Opus 5.5, GPT-6, Gemini 3.8, Grok 4.7 and other current entries; removed simulated benchmark fluctuations.
- Tool directory: removed synthetic ratings, initial vote counts and invented first-person product claims. Every current profile now includes official documentation and practical limitations.
- Learning and Lab: replaced empty course promises and imaginary project histories with available reading lessons, exercises, working calculators and downloadable worksheets.
- Media and editions: replaced nonexistent videos and fake runtimes with official resources and viewing briefs. The Signal editions are readable web publications; email delivery is not claimed.