おはようございます

モデルリリース

qwen3.8-omni-flashQwen

• 発信元: Alibaba Qwen / Tier1 / 日付: 2026-09-18 / 確度: 0.82
• 要約: happyoyster-1.0-adventureHappyOyster Adventure is a real-time interactive, explorable open-world model. It generates world scenes from multimodal (text/image) input, enabling immersive, boundless exploration through real-time interaction, movement, and camera control. Suitable fo
• 出典: https://docs.qwencloud.com/changelog/models#2026-09-18-happyoyster-1-0-adventurehappyoyster-adventure-is-a-real-time-interac

qwen3.8-livetranslate-flash-realtimeQwen3.8-LiveTranslate-Flash-Realtime

• 発信元: Alibaba Qwen / Tier1 / 日付: 2026-09-17 / 確度: 0.82
• 要約: qwen3.8-livetranslate-flash-realtimeQwen3.8-LiveTranslate-Flash-Realtime is a multilingual real-time audio/video simultaneous translation model. It supports audio and image input, text and audio output, recognizes 60 source languages, and supports voice output in 29 languages. Co
• 出典: https://docs.qwencloud.com/changelog/models#2026-09-17-qwen3-8-livetranslate-flash-realtimeqwen3-8-livetranslate-flash-realt

API・開発者向け更新

 The Compliance API local session endpoints now also return transcripts of Claude in Chrome sessions (product_surface

• 発信元: Anthropic / Tier1 / 日付: 2026-09-18 / 確度: 0.82
• 要約:  The Compliance API local session endpoints now also return transcripts of Claude in Chrome sessions (product_surface value claude_in_chrome), in beta for Claude Enterprise organizations, with your existing Compliance Access Key and the read:compliance_user_data scope. See Sessi
• 出典: https://platform.claude.com/docs/en/release-notes/overview#2026-09-18-the-compliance-api-local-session-endpoints-now-also-return-transcript

New back-to-school features and learning tools available in Gemini Notebook

• 発信元: Google Workspace / Tier1 / 日付: 2026-09-18 / 確度: 0.82
• 要約: We’re introducing several updates to Gemini Notebook that give users a more personalized and powerful learning experience.Audio recorder: Users can now use a new audio recorder in the Gemini Notebook mobile app (Android/iOS) to capture lectures or record thoughts on the go, with
• 出典: http://workspaceupdates.googleblog.com/2026/09/new-back-to-school-features-and-learning-tools-available-in-Gemini-Notebook.html

Antigravity Agent 09-2026: Released antigravity-preview-09-2026, which replaces and deprecates antigravity-preview-05-2026

• 発信元: Google AI for Developers / Tier1 / 日付: 2026-09-17 / 確度: 0.82
• 要約: Antigravity Agent 09-2026: Released antigravity-preview-09-2026, which replaces and deprecates antigravity-preview-05-2026. If you run on a remote sandbox (environment: "remote") and read only output_text or model_output steps, update the agent string and nothing else changes. If
• 出典: https://ai.google.dev/gemini-api/docs/changelog?hl=en#2026-09-17-antigravity-agent-09-2026-released-antigravity-preview-09-2026-which-

Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking generally available (GA): Released two new audio-to-audio models

• 発信元: Google AI for Developers / Tier1 / 日付: 2026-09-15 / 確度: 0.82
• 要約: Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking generally available (GA): Released two new audio-to-audio models for real-time voice applications using the Live API: Gemini 3.8 Live (gemini-3.8-live): The default option for most low-latency voice agent experiences and real-
• 出典: https://ai.google.dev/gemini-api/docs/changelog?hl=en#2026-09-15-gemini-3-8-live-and-gemini-3-8-live-extended-thinking-generally-avail

 The Messages API can now compact a conversation on demand on the Claude API, in beta with the compact-2026-09-04 beta header

• 発信元: Anthropic / Tier1 / 日付: 2026-09-14 / 確度: 0.82
• 要約:  The Messages API can now compact a conversation on demand on the Claude API, in beta with the compact-2026-09-04 beta header. Send the top-level compaction parameter, and the API returns a signed compaction block that summarizes the messages you sent. On later requests, send th
• 出典: https://platform.claude.com/docs/en/release-notes/overview#2026-09-14-the-messages-api-can-now-compact-a-conversation-on-demand-on-the-clau

Claude Code v2.1.278

• 発信元: Anthropic / Tier1 / 日付: 2026-09-19 / 確度: 0.82
• 要約: What's changed Changed auto mode for Claude API and Enterprise users, and on Bedrock, Vertex, Foundry and gateways, to default to the server-side classifier, which does not charge for classifier overhead (CLAUDE CODE AUTO MODE SERVER=0 opts out on Bedrock, Vertex, Foundry and gat
• 出典: https://github.com/anthropics/claude-code/releases/tag/v2.1.278

変化の読み方

• 今週のAI系ツールの進化は「API・開発者向け更新」が6件で最多（全8件）。
• 発信元は4系統。最多はAnthropicの3件で、単一ソース依存の度合いはここで判断する。
• 一次情報（Tier1）は8件、二次情報（Tier2/3）は0件。