Search covers subjects, readings, releases, communities, measured statements, quotes as pages show them, the glossary and the method. It never searches posts: the posts behind each figure are its fixed sample, on its page.

Release 2026-W23 · 1–7 Jun 2026 · published

Against the previous reading

Each subject against its previous stored reading under the same method. Each figure opens on its own reading page; the release page has the week’s summary table. The same comparison for share of posts is on its own page.

2026-W23 against the previous reading: people posting on each platform

Each subject against its previous stored reading under the same method, platform by platform. Each cell shows the earlier week’s figure, then this week’s; platforms are never added together.

No subject has a comparison to show in 2026-W23.

97 subjects have no earlier reading to compare

The week before, release 2026-W22, was read with different filters, so no change is shown: Alibaba, Anthropic, AutoClaw, Bard, ChatGLM, ChatGPT, Claude, Claude API, Claude app, Claude Code, Claude Cowork, Codestral, Codex, DALL·E, DeepSeek, DeepSeek, DeepSeek API, DeepSeek app, DeepSeek Harness, Devstral, Gemini, Gemini API (Google AI Studio), Gemini app, Gemini CLI, Gemini Spark, Gemma, GitHub Copilot, GLM, AI Mode, AI Overviews, Google Antigravity, Google DeepMind, GPT, GPT Image, GPT-Live, gpt-oss, Grok, Grok app, Grok Build, Hailuo, Hailuo AI, Jules, Kimi, Kimi API Platform, Kimi app, Kimi Code, La Plateforme, Le Chat, Leanstral, Llama, Magistral, MAI, Meta, Meta AI, Microsoft AI, Microsoft Copilot, Microsoft Foundry, MiniMax, MiniMax-01, MiniMax Agent (Mavis), MiniMax Open Platform (API), MiniMax Code, MiniMax H-series, MiniMax M-series, Ministral, Mistral, Mistral AI, Mistral Vibe, Mixtral, Moonshot AI, Muse, Muse (agent app), Muse Code, NotebookLM, OpenAI o-series, OpenAI, OpenAI API, Phi, Pixtral, Qwen, Qwen App, Qwen Code, Qwen Studio, QwenCloud, QwQ, Sora, Sora app, Talkie, Veo, Voxtral, Wan, Whisper, xAI, xAI API, Z.ai API Platform, ZCode, Zhipu AI.

Population: each subject’s posts on each platform, published in 2026-W23 and in the week of its previous stored reading under the same method. n not recorded; each cell’s n is under its figures, the earlier week’s then this week’s, and each week’s population, coverage and, for an estimate, its 95% interval below that. The page doesn’t say whether a figure rose or fell: that’s for AIDRAN’s stored change rule, which hasn’t been run on these readings yet.