OpenCode v1.17.6 → v2.3.0: what changed
65 releases shipped between v1.17.6 (June 13, 2026) and v2.3.0 (November 13, 2026).
Changelog entries in this range
- v1.17.7 — June 14, 2026: Core Bugfixes - Plugin client requests now reuse the active server instead of assuming the default local port. - ACP shell tool calls now show the command and w
- v1.17.8 — June 17, 2026: Core Improvements - Session timelines load much faster and avoid flicker or scroll jumps. Bugfixes - OpenAI-compatible providers now accept MCP tool schemas tha
- v1.17.9 — June 21, 2026: Core Bugfixes - Honor configured agent step limits by forcing a final text response instead of failing mid-run. - Fix Devstral model detection when provider IDs
- v1.17.10 — June 24, 2026: Core Improvements - Added MCP server instructions to session context. (@Arcadi4) - Added Opencode-managed provider integration support. - Added MCP resource tem
- v1.17.11 — June 25, 2026: Core Improvements - Add session snapshots and revert controls so you can roll a session back to an earlier message, including file changes. Bugfixes - Always pr
- v1.17.12 — June 30, 2026: Core Bugfixes - Enable adaptive thinking for Claude Sonnet 5. - Prefer MCP content responses over structured output when both are present. - Reconnect MCP serve
- v1.17.13 — July 1, 2026: Core Bugfixes - Force reasoning mode for OpenAI-compatible reasoning models so reasoning settings apply reliably on custom deployments - Stop replaying stale Gi
- v1.17.14 — July 6, 2026: Core Improvements - Added a code mode MCP adapter for running confined orchestration scripts against connected MCP tools. - Hid the execute tool unless code mod
- v1.17.15 — July 7, 2026: Core Bugfixes - Better classify Z.ai context-window overflow errors so oversized requests surface the right failure mode (@fengjikui) - Handle unavailable confi
- v1.17.16 — July 9, 2026: Core Bugfixes - Exposed reasoning effort variants for Grok models. - Improved xAI prompt cache routing and PDF file support in Responses models. Desktop Improve
- v1.17.17 — July 9, 2026: Core Bugfixes - Improved Meta model handling for reasoning variants and provider requests. Desktop Bugfixes - Fixed clipped descenders in model selector labels.
- v1.17.18 — July 9, 2026: Core Bugfixes - Prevent crashes and bad pricing data when GitHub Copilot returns models with a zero billing batch size. Improvements - Add a model-specific syst
- v1.17.19 — July 13, 2026: Core Bugfixes - Supported OpenAI pro reasoning mode. - Disabled response storage by default for xAI Responses. (@geraint0923) - Added OAuth support for Luna Res
- v1.17.20 — July 13, 2026: Core Bugfixes - Removed an obsolete Codex workaround that could interfere with OpenAI Luna Responses Lite requests. Improvements - Updated Azure AI support for
- v1.18.0 — July 14, 2026: Desktop Improvements - Completed the Desktop v2 migration, including upgrade handling for the new layout and first-launch onboarding. - Added a setting to switc
- v1.18.1 — July 14, 2026: Desktop Bugfixes - Fixed spacing between model provider sections in Settings.
- v1.18.2 — July 15, 2026: Core Bugfixes - Stopped subagents from launching nested subagents by default, with a configurable subagentdepth limit when needed. - Improved default reasoning
- v1.18.3 — July 16, 2026: Core Improvements - Added an Up Arrow shortcut to close the subagent picker when the first item is selected. Desktop Bugfixes - Fixed home page scrolling so sti
- v1.18.4 — July 20, 2026: Core Improvements - Use adaptive thinking controls for Kimi models on Anthropic-compatible providers, with summarized reasoning output by default. (@chouqin) Bu
- v1.18.5 — July 24, 2026: Core Bugfixes - Improve Claude adaptive thinking handling across more response shapes. - Avoid OpenAI Responses phase handling that could break some conversatio
- v1.18.6 — July 27, 2026: Core Bugfixes - Fixed branch-specific repository caches so refreshing one reference no longer moves another branch checkout. Desktop Improvements - Improved com
- v1.18.7 — July 27, 2026: Desktop Bugfixes - Remove the extra titlebar inset in fullscreen on macOS. - Keep command palette entries from reappearing incorrectly when shadowed commands ar
- v1.18.8 — July 28, 2026: Core Improvements - Improved compatibility with newer MCP servers and OAuth flows. Bugfixes - Reconnects MCP servers after expired SDK sessions, including concu
- v1.18.9 — July 28, 2026: Core Bugfixes - Restored compatibility with legacy MCP SDK clients. Desktop Bugfixes - Fixed a Solid cleanup crash that could break navigation in the desktop ap
- v1.18.10 — July 30, 2026: Core - Discover available Modal models automatically. (@devennavani) Desktop Improvements - Prevent adding the same attachment more than once. - Always show the
- v1.18.11 — August 1, 2026: Core Bugfixes - Stopped MCP SSE connections from getting stuck in reconnect loops after server error responses. - Fixed provider model configs that use interlea
- v1.18.12 — August 4, 2026: Core Bugfixes - Fixed Azure GPT-5.5+ completion requests failing when reasoning is enabled. (@frederiknsgo) Desktop Bugfixes - Reduced composer lag when drafts
- v1.18.13 — August 4, 2026: TUI Bugfixes - GitHub pull request reviews now include the pull request number and URL in context. Desktop Bugfixes - Fixed several right-to-left layout issues
- v1.18.14 — August 5, 2026: Core Improvements - Simplified xAI login to a single device-code flow that works better in headless and remote environments. Bugfixes - Preserved structured mid
- v1.18.15 — August 7, 2026: Core Bugfixes - Chronological message ordering now stays correct even when imported or legacy message IDs are out of order. - Revert and fork actions now use re
- v1.18.16 — August 10, 2026: Core Bugfixes - Ignore unknown top-level config fields instead of failing config parsing - Register projects opened from Home so they are available to the rest
- v1.18.17 — August 12, 2026: Core Bugfixes - Made session compaction keep complete recent turns and produce clearer summaries for smaller models. - Added MERGE Gateway reasoning variants so
- v1.18.18 — August 13, 2026: Core Bugfixes - Select the Kimi system prompt correctly for official Moonshot and Kimi providers - Fix xhigh reasoning effort for xai models
- v1.18.19 — August 20, 2026: Core Improvements - Added native OpenAI and Anthropic passthroughs for Cloudflare AI Gateway models. - Matched Codex rate limits more closely to ChatGPT subscri
- v1.18.20 — August 21, 2026: Core Bugfixes - Surface failed subagent tool calls with a resumable taskid. - Retry provider responses that end with finishreason: networkerror. - Retry more ne
- v1.18.21 — August 21, 2026: Core Bugfixes - Continue responses when a model reports an unknown finish reason instead of stopping early - Route Vertex AI eu and us multi-region Gemini reque
- v1.18.22 — August 24, 2026: Core Bugfixes - Removed outdated OpenCode Go first-month discount messaging and pricing. - Fixed OpenCode device login links when servers return relative verifi
- v1.18.23 — August 25, 2026: Core Bugfixes - Fixed Cloudflare AI Gateway routing for third-party providers so non-Workers models work through the gateway's REST API. (@superhighfives) - Fix
- v1.18.24 — August 28, 2026: Core Bugfixes - Bedrock reasoning responses no longer get cached into unreplayable empty messages. Improvements - Azure providers can now sign in with Microsoft
- v1.18.25 — August 28, 2026: Core Bugfixes - Fixed Azure authentication so Azure CLI sign-in works without requiring Bun.
- v1.18.26 — September 1, 2026: Core Bugfixes - Claude 5 sessions now tolerate stale thinking blocks instead of failing after prompt or tool changes. - Bedrock GPT-5.6 models now accept none r
- v1.18.27 — September 2, 2026: Core Bugfixes - Default provider header timeouts to five minutes so slow model startups fail less often. - Default streamed chunk timeouts to five minutes, with
- v1.18.28 — September 4, 2026: Core Improvements - Send the session ID as GitHub Copilot's interaction header to improve request tracking across a session. Desktop Bugfixes - Use the desktop
- v1.18.29 — September 4, 2026: Core Bugfixes - Allow Codex OAuth model filtering to recognize integer GPT versions like gpt-6 - Fixes issue of gpt-6-astra not showing up for openai subscripti
- v1.18.30 — September 9, 2026: Core Improvements - Added the Astra system prompt for GPT-6 models. Bugfixes - Preserved Bedrock DeepSeek model IDs, including ARN-based IDs, so they resolve co
- v1.18.31 — September 14, 2026: Core Bugfixes - Restored ACP session model, effort, mode, and reasoning chunk boundaries when loading, resuming, or forking sessions. (@JacobNWolf) TUI Bugfixes
- v1.18.32 — September 21, 2026: Core Bugfixes - Fixed Bedrock image attachments so they are only hoisted for Claude, Nova, and Llama 4 models. - Fixed Together AI streaming usage reporting. Th
- v1.18.33 — September 28, 2026: Core Bugfixes - Cloudflare AI Gateway models now honor provider response and stream timeouts. (@danlapid) - MCP browser launch failures are now reported when th
- v1.18.34 — September 30, 2026: Core Bugfixes - Send namespaced session and parent-session identity headers with model requests. - Re-sign locally compiled macOS binaries so they run reliably
- v1.18.35 — October 6, 2026: Core Improvements - Added canonical redirects and JSON and Markdown data formats for agent-readable stats. Bugfixes - xAI tool results now include supported ima
- v1.19.0 — October 11, 2026: Core Improvements - Run opencode upgrade --major to upgrade to OpenCode 2. Bugfixes - Allow the OpenAI ultrafast service tier. - Validate PTY connection origins
- v1.19.1 — October 15, 2026: Core Bugfixes - Session sharing has been discontinued. Existing share links and session-sharing commands are no longer available, and importing sessions from sh
- v1.19.2 — October 16, 2026: Core Bugfixes - OpenCode 1 now stays on the 1.x release line: update checks use only the npm opencode-ai latest version, so 1.x installs are never offered a 2.x
- v1.19.3 — October 16, 2026: Core Improvements - Configure model reasoning options to generate reasoning variants, replacing catalog variants while preserving your configured variant overri
- v1.19.4 — October 22, 2026: Core Improvements - Upgraded the bundled Bun runtime to 1.4.4, which includes security fixes for Brotli, TLS, HTTP/1.1, HTTP/2, HTTP/3, WebSocket, and fetch, an
- v2.2.0 — October 27, 2026: OpenCode 2.2.0 is here: build and share Desktop plugins, plus new keyed session instructions that keep your prompt cache intact. Breaking changes inside, so che
- v1.19.5 — October 28, 2026: Ever had OpenCode freeze when a TUI worker request failed? In 1.19.5, failed requests now return a proper error instead of hanging. Thanks to @Nowaker for the f
- v2.2.1 — October 29, 2026: OpenCode 2.2.1 lets your session keep going until a reviewer confirms the goal is met. Plus image generation, a smarter model picker, and live background shells
- v2.2.2 — October 29, 2026: OpenCode 2.2.2 is here: image generation works out of the box with supported providers, the Git branch on new sessions opens your worktree picker in one click,
- v2.2.3 — October 30, 2026: OpenCode 2.2.3 lets you use your Claude Code login and models right inside OpenCode, and long Desktop sessions now run smoother. Breaking changes hit @opencode/
- v2.2.4 — October 30, 2026: Desktop terminals now survive restarts with scrollback intact, and agents can publish HTML artifacts. Claude Agent setup got a guided walkthrough too. Curious w
- v2.2.5 — October 30, 2026: OpenCode 2.2.5 is here: percentage compaction buffers, one-key worktree switching, and send-quote shortcuts. Shell calls now just work too. Curious what else ch
- v2.2.6 — November 6, 2026: Share one MCP connection across projects, run scripted sessions in fresh git worktrees, and keep CI runs out of your inbox. OpenCode 2.2.6 is here. Want to see
- v2.2.7 — November 10, 2026: Let Copilot pick the model for you with Auto, close side panel tabs faster, and mark sessions unread so they stay flagged. OpenCode 2.2.7 is out. What else did
- v2.3.0 — November 13, 2026: OpenCode 2.3.0 is here: subagents now run quietly in the background while you keep working, transcripts collapse into tidy activity rows, and you can jump to an