OpenAI's Oct 16 release notes say Slides lets you build decks from a chat, files, or an outline in ChatGPT Work, co-edit them in Space, and export to PowerPoint, PDF, or Google Slides on the web. GPT-6 Pro in Chat also now splits a task across multiple agents running in parallel, each with a viewable record of what it did, rolling out to eligible Pro users.
As of Oct 16, adding https://api.heypcb.ai/mcp to Claude Code, Codex, Cursor, or ChatGPT lets agents draw schematics, place parts, route boards, run checks, export KiCad projects, and design enclosures. In HeyPCB's demo, Claude Code laid out a 145 part 8x8 RGB display board, fixed a rotated LED that wouldn't route, and finished with zero DRC and ERC errors in KiCad…
As of Oct 16, Cursor cloud agents can produce interactive canvases as part of their work, shareable with your team by link, and a beta Environments API lets you create and manage the cloud environments agents run in from code. Those environments now hold up to 300 repositories, and large ones load 5x faster…
As of Oct 16, any API customer can describe a voice in text via POST /v1/audio/voices or the voice Playground (free to create, up to 100 per org) and use it in gpt live 1 sessions at normal session rates; these voices don't work with TTS, Realtime, or Chat Completions. GPT Live also gains nine natural German, French, Hindi, Japanese, and Korean voices…
OpenCode's model list now includes glm 5.3 flash free and beam free (select them as opencode/<id ), both priced at zero for input, output, and cached tokens with no end date announced; the paid GLM 5.3 Flash stays at $0.15/$0.50 per 1M tokens. The GLM provider keeps no data, while Beam Free retains prompts for 30 days for safety and is blocked in some countries, including China and Russia…
As of Oct 16, adding https://mcp.teable.ai/mcp to an MCP client (for example claude mcp add transport http teable https://mcp.teable.ai/mcp ) lets agents list projects, inspect tables and fields, query and edit records, and discover APIs within your existing permissions, approved once in the browser instead of with a personal access token. Self hosted instances expose it at /mcp , and Settings → MCP lists connected clients so you can revoke them…
Spotted something we missed? Start a thread.
