Released: 2026-07-21
Long conversations no longer have to feel like one endless scroll. Experimental paginated history lets you resume efficiently, search, keep names, and work with sub-agents and memories. Think of a loose stack of meeting notes moved into an indexed binder: find the page, then continue from it. Experimental means it may still change.
Moving to Codex no longer means rebuilding every old habit. /import can migrate Cursor and Claude Code settings, MCP servers, plugins, sessions, commands, and project-scoped memories. It is more like moving a whole phone than just its contacts. Run /import, then choose the source and what to bring over; compatibility is not promised for every setting.
Teams using Amazon Bedrock now get experimental login, custom endpoint, and authentication support, with GPT-5.6 Sol as the default Bedrock model. It is like adding a badge entrance and allowing a company’s own phone line into the building. This matters mainly when Codex must fit an existing Bedrock setup.
When the source is a recording rather than text, Codex can now take common local audio formats as input and return audio from tools. Streaming realtime V3 conversations also arrive: replies begin traveling while they are generated, like a phone call instead of a finished voicemail. Exact formats and commands are not specified here.
For work split across several agents, the opt-in multi-agent V2 experience is now more stable. You can configure sub-agent models, reasoning levels, and concurrency; roles return and navigation improves. It is like a director choosing each crew member’s gear and number of simultaneous tasks. Opt-in means it must be enabled deliberately.
Visualizations in terminal output can now include secure, clickable inline links. Instead of copying an address and guessing where to paste it, you can open the full view from the result itself. The release only promises secure clickable links, not universal support for every terminal or visualization.
Editing an earlier prompt or retrying a safety-buffered turn now makes a contextual branch. The original conversation, attachments, and mention bindings stay intact. It is like photocopying a manuscript before editing: the new draft can diverge while the original and its materials remain available.
Long terminal conversations now redraw less and feel more responsive. Incremental Markdown rendering, fewer redraws, caching, and bounded command output improve long chats and streamed output. It is like a cashier adding one line to a receipt rather than reprinting the whole thing after every item. Output limits also keep one command from flooding the screen.
When many external tools start at once, MCP sign-ins and credential refreshes can collide. Startup timeouts, non-blocking OAuth discovery, serialized refreshes, and safe reuse of tool catalogs prevent slow or conflicting flows. Think of an airport that keeps one delayed passenger from blocking every gate and reuses one flight board instead of printing it repeatedly.
Windows execution and sandbox reliability are strengthened with native exec-server sandboxing, enforced network-proxy rules, hidden helper consoles, and correctly quoted hook commands. It is like fixing a worksite’s fence, entrance, and radio protocol together: invisible details that keep the job inside its boundaries and running as intended.
Installation should not stop merely because an external host is temporarily unavailable. Compact release-metadata parsing and macOS code-mode installation are fixed, with an in-process fallback when the external code-mode host cannot run. Like an automatic door retaining an internal button when its outside sensor fails, this is a repair and fallback for the existing path.
Safety handling now detects forced rm deletion more carefully, confirms full access more consistently, and preserves rejection reasons across tools. It is like a guard who notices not only a key but also an attempt to remove the whole door, then hands your refusal note to the next guard. This strengthens protection; it does not loosen deletion rights.
Bundled OpenAI Docs guidance now covers GPT-5.6 model resolution, prompting, and migration across macOS, Linux, and Windows. Packaged GPT-5.4 selections and internal uses also move to the corresponding GPT-5.6 Terra and Luna variants. It is like updating both the catalog and shelf labels so model choices reflect the current generation.
Startup and large-context work get lighter through concurrent skill and plugin discovery plus more efficient remote compaction. The bundled ripgrep, a fast text-search tool, also moves to 15.2.0. It is like a kitchen preparing ingredients in parallel instead of one station waiting on the next, then upgrading its fastest finder.
Added experimental paginated thread history with efficient resume, search, persisted names, sub-agent support, and memories. (#33364, #33907, #34085, #34229, #34386)Expanded `/import` to migrate Cursor and Claude Code settings, MCP servers, plugins, sessions, commands, and project-scoped memories. (#31672, #33411, #33426, #33444)Added experimental Amazon Bedrock login, custom endpoint and authentication support, and GPT-5.6 Sol as the default Bedrock model. (#31327, #33170, #33175, #32288, #33695)Added audio inputs and tool outputs, including common local audio formats, and introduced streaming realtime V3 conversations. (#33261, #33856, #33932, #34080, #34385)Stabilized the opt-in multi-agent V2 experience with configurable sub-agent models, reasoning levels, concurrency, restored roles, and improved agent navigation. (#33550, #33631, #33657, #33841, #34383)Added secure, clickable inline visualization links in the terminal UI. (#33925, #34217, #34346)Editing an earlier prompt or retrying a safety-buffered turn now creates a contextual branch, preserving the original conversation, attachments, and mention bindings. (#33201, #33207, #33211)Improved terminal responsiveness for long conversations and streamed output through incremental Markdown rendering, fewer redraws, caching, and bounded command output. (#34045, #34049, #34216, #34223, #34359)Prevented slow or conflicting MCP startup and authentication flows by enforcing startup timeouts, avoiding blocking OAuth discovery, serializing refreshes, and reusing tool catalogs safely. (#32229, #32781, #32825, #33184, #33297)Improved Windows execution and sandbox reliability, including native exec-server sandboxing, network-proxy enforcement, hidden helper consoles, and correctly quoted hook commands. (#32849, #32857, #33926, #34423)Fixed compact release-metadata parsing and macOS code-mode installation, with an in-process fallback when the external code-mode host is unavailable. (#31667, #31876, #31899)Strengthened safety and approval handling with better forced-`rm` detection, consistent full-access confirmation, and preserved rejection reasons across tools. (#32989, #33464, #34400)Updated the bundled OpenAI Docs skill with current GPT-5.6 model resolution, prompting, and migration guidance across macOS, Linux, and Windows. (#31842, #33121)
Migrated bundled GPT-5.4 selections and internal uses to the corresponding GPT-5.6 Terra and Luna variants. (#33173)Reduced startup and large-context overhead with concurrent skill/plugin discovery and more efficient remote compaction. (#31566, #33369, #33423, #34431)
Updated the packaged ripgrep binary to 15.2.0. (#34384)