cherry-studio v2.2.3
cherry-studio v2.2.3
Downloads (v2.2.3)
| Platform | Architecture | Download |
|---|---|---|
| Windows | x64 | Installer · Portable |
| Windows | arm64 | Installer · Portable |
| macOS | Apple silicon (arm64) | DMG · ZIP |
| macOS | Intel (x64) | DMG · ZIP |
| Linux | x64 | AppImage · DEB · RPM |
| Linux | arm64 | AppImage · DEB · RPM |
Release Notes
Cherry Studio 2.2.3 - Conversation Sorting, New Models, and Agent Reliability
✨ New Features
- [Conversations] Grouping and sorting are now independent. Sort by last updated, creation date, name (with natural number order), or manual order, and existing custom orders are preserved.
- [Agent] Import DeepSeek v3 exports into an automatically created DeepSeek Agent, keeping history, branches, reasoning, search records, and available attachments.
- [Agent] A short next prompt is suggested after replies: press Tab to accept without sending, or Escape to dismiss. Suggestions can be turned off in input settings.
- [Agent] You can upload an image avatar while creating an Agent.
- [Agent] Automatic command approval now parses Bash and PowerShell with shared risk rules and asks before each risky command. Approval cards show the requesting task, working directory, and reason. (action required: review permission modes if you rely on the existing prompts; unmatched commands run automatically, while mandatory denials still apply in Full Access.)
- [Browser] Conversation sidebars support multiple retained tabs with unified styling and context menus, use the configured homepage, and close with their last page.
- [File Preview] PPTD slide decks now preview in the app with page navigation, zoom, thumbnails, refresh, and a source view.
- [Files] File paths and Markdown links with line numbers jump to and highlight the referenced line in the preview.
- [Files] The file selection dialog opens in the folder of the last file you selected during the current session.
- [Chat] Added an Appearance → Messages preference, “Scroll to the latest message on send” (on by default).
- [Chat] Long deliverable file lists show the first three files and let you expand or collapse the rest.
- [Chat] Ctrl + mouse wheel zooms the interface, shows the zoom level while scrolling, and Ctrl+0 resets it.
- [Knowledge] The URL source accepts several links at once (separated by newlines, commas, or spaces), adds each valid URL separately, and reports invalid entries individually.
- [Paintings] Generated images show the model and returned size, and are saved in their original format instead of always being converted to PNG.
- [Computer Use] A new Linux-only “Global input” switch (off by default) lets agents control the shared mouse pointer and system keyboard.
- [Models] Added GPT-6.2 Astra and Sol, Gemini 4 Carbon and Hydrogen with reasoning controls, Tencent Hunyuan 4 Flash, and Mistral Large 4 reasoning controls.
- [Models] DeepSeek first-party cost estimates now use weekday peak and off-peak rates while keeping custom prices. (action required: review local cost estimates if you relied on peak-only estimates.)
- [Models] Enabled providers now appear first in the model service list, and primary model pickers can open in a remembered dialog mode.
- [Marketplace] Curated skills can be filtered by category, with search and horizontal category navigation.
- [Composer] Pasting copied folders into Agent conversations adds folder references instead of failing file attachments.
- [Export] Obsidian exports label each assistant message with the model that produced it.
- [Shortcuts] Number shortcuts (1–9) have an explicit Off preset in Settings → Shortcuts.
- [Voice] The recognition model now has an Off option that disables local dictation without removing the saved language or hotkeys.
- [Code CLI] Qwen Code 1.0 can be installed from its GitHub release.
- [App] Added Czech (cs-CZ) language support.
🐛 Bug Fixes
- [Chat] Nested scrollable content such as code previews, editors, and embedded artifacts scrolls independently while the message list scrolls.
- [Chat] Conversations restored from Settings → Archive open with their messages instead of an empty conversation.
- [Chat] Interrupted streamed responses are reported as connection errors with partial output kept for retry, instead of an invalid AI response.
- [Chat] Multi-model fold tabs show every model name again, with a toggle back to compact tabs.
- [Chat] Chat requests that exceed the model context window recover more reliably.
- [Notes] LaTeX written with [ … ] and ( … ) renders again in the Notes editor.
- [Chat] The browser side-panel resize bar is draggable along its full height.
- [Agent] Assistant text emitted before or between tool calls is visible again in Agent sessions.
- [Agent] Pi Durable: Stop keeps the session open and the reply paused; reopening after quitting continues an interrupted reply or shows its real result; messages sent while the agent works are saved and answered in order; a withdrawn message is marked “Stopped before sending” with Resend and Edit.
- [Agent] Messages sent while a turn is running are acknowledged and acted on instead of being silently ignored.
- [Agent] A pending approval is cleared when its background task ends, so the reply no longer stays on Pending.
- [Agent] Model selection is scoped to each session, so changing one session no longer changes sibling sessions.
- [Agent] Claude Code conversations can be forked or edited after tool retries or compaction when the SDK can preserve their context.
- [Agent] In Accept Edits mode, Claude Code asks before running an escalated Bash command instead of auto-running it by its first word.
- [Agent] Agent session history repairs message trees during upgrade and prevents edits to in-flight deliveries.
- [Agent] An Agent turn recovers when a directory occupies its durable store name instead of failing with a network error.
- [Agent] Reasoning-effort, service-tier, and Fast controls stay available when editing and resending a historical message.
- [Agent] Historical messages follow the Agent’s current avatar, and an avatar chosen late during creation is applied.
- [Agent] Selection assistant: reused windows no longer open at full conversation size, a closed window no longer leaves a stray capsule, and the panel opens above the active app.
- [Agent] Agent requests delivered through HTTP or in-process gateways keep protocol metadata, Gemini configuration, generated files, and continuation state.
- [Agent] Unreadable stored automation tool results are reported to the model as a note instead of breaking the conversation.
- [Agent] NO_PROXY no longer includes bracketed IPv6 or IPv6 CIDR entries that broke Python HTTP clients in Agent sessions.
- [Models] Model connection checks now use streaming for streaming-capable models and detect incomplete or failed responses.
- [Models] Unsupported sampling parameters and penalties are no longer sent to providers, and declared sampling ranges apply after request overrides.
- [Models] Gemini tool calls no longer fail when a schema contains propertyNames.
- [Models] A single Cherry Cloud quota refusal no longer locks a model while plan quota remains; the server decides refusals and shows its suggested wait.
- [Models] CherryIN saved model endpoints refresh automatically from the server while keeping local values when discovery fails.
- [Models] CherryAI appears first with a paid subscription and last without one.
- [Models] Provider request rejections without an HTTP status are reported as request errors instead of network interruptions.
- [Models] Usage stats recover provider cache buckets, and locally computed costs are labeled as estimates.
- [Models] Agent sessions on OpenAI-compatible, vLLM, or llama.cpp models report how far a request exceeds the context window.
- [Models] Aborted provider turns now fail instead of reporting a successful stop.
- [Files] File type icons render offline and stay consistent across file trees, previews, and attachments.
- [Files] File trees and previews stay current, show save progress and failures, and support scoped refresh without losing local drafts.
- [Knowledge] The Add URL dialog rejects non-http(s) addresses immediately.
- [Knowledge] Self-hosted MinerU 4.1 Tasks API parsing is supported while older protocols still work.
- [Feedback] Problem reports again include logs, activity records, and chat history for the last 24 hours; individual attachment switches were removed.
- [Scheduled Tasks] Run history recorded before 2.2.0, or whose conversation was deleted, can be opened and shows its stored result.
- [Skills] Cherry-only built-in skills are no longer exported to the shared ~/.agents/skills directory.
- [Feishu] Initial channel connections retry with backoff instead of staying stopped after a failed handshake.
- [UI] Empty assistant, agent, and notes sidebars explain that they are empty instead of showing a blank panel.
- [Agent] Fixed agent sessions with background subagents slowing down a little more every round.
- [Agents] Preserve subagent rate-limit error details and show consistent retry status for Pi and Pi Durable tasks.
- [Agents] Reduce out-of-memory crashes when subagents process many images in parallel.
- [Cherry AI] Load the model catalog from the server and show sign-in, subscription or AMD authorization requirements; detect delisted models. Built-in Qwen now requires Cherry AI sign-in (action required for users who previously used it signed out).
💄 Improvements
- [Conversations] Grouping choices are clearer, and list controls are separated from history and management navigation in conversation and Agent session menus.
- [Agent] The service tier picker explains each tier’s trade-offs and shows when a saved tier is unavailable, with provider-specific names in a collapsible details view.
- [Agent workspace] The folder “Open with” menu now lists every application registered on the system for opening folders, with system icons, instead of only the built-in editors and terminal.
- [UI] Brand icons, mini-app labels, settings layouts, help, skill controls, and marketplace visuals are more consistent.
- [UI] Launchpad and Mini Apps icons use consistent sizing and corners with refreshed dark-mode artwork, and inset scrollbars in file trees and lists align with pane edges.
- [Diagnostics] Native crash reports upload automatically to Sentry when data collection is enabled.
- [Data upgrade] v1 users must migrate through Cherry Studio 2.2.2; eligible installations offer one-click installation of the migration release, while others keep direct platform-matched downloads. (action required)
⚡ Performance
- [Agent] Agent sessions with many images or subagents use much less main-process memory and CPU; finished subagents are released and Durable history is read incrementally.
- [Agent] Restoring large Claude Code sessions and loading their history uses less main-process memory.
- [Chat] Streaming responses and conversation history loading use fewer copies of conversation snapshots.
What’s Changed
- fix(external-app): list every registered folder application in open with by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24348
- fix(release-cos): preserve Windows regional uploads with architecture arguments by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24446
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24449
- fix(release): isolate Windows COS upload dependencies by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24451
- chore(release): sync v2.2.2 metadata by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24452
- fix(skills): stop exporting Cherry-only builtins to ~/.agents/skills by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24450
- docs(chat-context): propose overflow recovery by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24354
- fix(chat-context): recover from context overflow by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24410
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24456
- fix(agent): clear pending approval when a background task ends by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24444
- fix(windows-signing): avoid duplicate SHA-256 signing passes by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24457
- perf(release-pipeline): overlap draft uploads and edition packaging by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24455
- perf(release-tools): cache and overlap external tool downloads by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24458
- perf(packaging): trim duplicate SQLite and Tesseract binaries by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24462
- feat(marketplace): filter curated skills by category by @ousugo in https://github.com/CherryHQ/cherry-studio/pull/24466
- fix(scroll-area): align file tree and list scrollbars with pane edges by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24467
- feat(session-browser): add retained sidebar browser tabs by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24468
- fix(ui): preserve focus for portaled input-group controls by @ousugo in https://github.com/CherryHQ/cherry-studio/pull/24474
- fix(agent): require an acknowledgement for mid-turn steer messages by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24463
- refactor(data-migration): replace v1 conversion with native installer downloads by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/23998
- feat(chat-artifacts): collapse long deliverable file lists by @1lck in https://github.com/CherryHQ/cherry-studio/pull/24509
- fix(message-group): restore the fold tab display toggle by @Doekin in https://github.com/CherryHQ/cherry-studio/pull/24504
- fix(agent): keep the avatar file input mounted so a late selection still applies by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24486
- feat(provider-settings): list enabled providers first in the model service list by @404-Page-Found with @Copilot in https://github.com/CherryHQ/cherry-studio/pull/23570
- feat(agent-composer): suggest the next prompt with Tab acceptance by @1lck in https://github.com/CherryHQ/cherry-studio/pull/24506
- feat(code-cli): install Qwen Code 1.0 from its GitHub release by @angri450 in https://github.com/CherryHQ/cherry-studio/pull/24510
- fix(claude-code): allow SDK-readable history to fork by @WhiteZeek in https://github.com/CherryHQ/cherry-studio/pull/24512
- feat(provider-registry): support Gemini 4 Hydrogen reasoning levels by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24517
- feat(models): add Gemini 4 Carbon reasoning effort controls by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24476
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24519
- fix(agent-sessions): scope model selection to the session instead of the agent by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/20826
- revert(feedback): restore diagnostic report attachments by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24532
- fix(scheduled-tasks): keep stored run results readable without a session by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24465
- feat(i18n): add Czech (cs-CZ) language by @markiceman in https://github.com/CherryHQ/cherry-studio/pull/24494
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24533
- fix(snapshot-projection): reduce copies and restrict structuredClone by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24535
- test(pi-durable): fix Windows child streaming fixture by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24544
- fix(claude-code): bound session archive restore memory by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24518
- feat(model-selector): add dialog mode to primary model pickers by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24529
- fix(i18n): translate the model picker mode toggle into Czech by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24557
- feat(agent-avatar): upload images during Agent creation by @ousugo in https://github.com/CherryHQ/cherry-studio/pull/24530
- docs(agent-notes): propose pptd slide preview spec by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24488
- docs(agent-notes): propose service tier explanations spec by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24501
- feat(provider-billing): add DeepSeek weekday pricing defaults by @WhiteZeek in https://github.com/CherryHQ/cherry-studio/pull/24527
- fix(cherry-cloud): order providers by subscription status by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24416
- fix(i18n): translate the provider billing period strings into Czech by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24589
- style(ui): unify settings, brand icons, and marketplace presentation by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24574
- fix(cherryin): refresh saved model endpoints from the server by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24531
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24584
- fix(sentry): upload native crash dumps automatically by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24573
- fix(feishu): retry initial channel connect with backoff by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/23376
- fix(extension-table-plus): point ESM exports and types at the emitted .mjs/.d.mts files by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24590
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24598
- fix(api-gateway): fail aborted provider turns instead of a successful stop by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24442
- fix(conversation-list): clarify grouping and navigation menus by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24599
- fix(build): extract Windows tool archives in-process by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24611
- refactor: remove exports with no remaining consumers by @angri450 in https://github.com/CherryHQ/cherry-studio/pull/24600
- feat(shortcuts): add an explicit Off preset for number shortcuts by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24617
- fix(error-category): classify status-less invalid_request_error as bad request by @zuhairkhalid229 in https://github.com/CherryHQ/cherry-studio/pull/24603
- test(e2e-regression): wait for the topic sidebar and knowledge import to settle by @cherry-qa[bot] in https://github.com/CherryHQ/cherry-studio/pull/24569
- fix(launchpad): align app icons and dark-mode artwork by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24615
- feat(chat): add a preference for scrolling to the latest message on send by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24597
- feat(computer-use): honor the user global input preference in SDK actions by @JeremySu0818 in https://github.com/CherryHQ/cherry-studio/pull/24595
- feat(product-guide): add in-app guide skill by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24542
- feat(voice): add an Off option to the recognition model to disable dictation by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24566
- fix(automation-repl): note unreadable history results instead of throwing by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/23306
- fix(agent-identity): send the identity snapshot inside one tag on Pi by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24621
- fix(voice-tests): repair main CI preference mock and clone lint by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24628
- fix(selection): keep the selection panel above the active app on every show by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24625
- fix(agent-session): enforce message tree integrity and improve traversal by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/23935
- fix(composer): paste copied directories as folder tokens by @ZKunZhang in https://github.com/CherryHQ/cherry-studio/pull/24629
- test(e2e): mock embeddings and remove CherryIN model scenarios by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24638
- ci(release): unify Tencent Cloud credentials by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24641
- fix(tool-approval): stop auto-approving acceptEdits Bash by first command word by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24639
- test(e2e): replace external test providers with local mocks by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24645
- fix(agent-session): read the deleted message once with only the needed columns by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24648
- fix(tool-schema): stop propertyNames reaching Gemini chat relays by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24664
- fix(tool-schema): guard the Gemini chat strip against partial scopes by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24667
- test(tool-schema): provide model in parameterless fixtures by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24666
- docs(web-search): propose StepFun text-to-image search spec by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24363
- docs(agent-notes): propose background system workspace archive by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24361
- docs(agent-notes): propose cherry-miniapp skill spec by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24359
- docs(agent-notes): propose session metrics bar spec by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24302
- fix(provider-settings): log model check timings and document check policy by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24130
- fix(api-gateway): classify SDK argument failures as request errors by @ZKunZhang in https://github.com/CherryHQ/cherry-studio/pull/24640
- fix(agent): keep historical message avatars current by @ousugo in https://github.com/CherryHQ/cherry-studio/pull/24478
- feat(file-storage): remember last directory in file selection dialog by @404-Page-Found with @Copilot in https://github.com/CherryHQ/cherry-studio/pull/23235
- Keep all assistant text visible in Agent sessions by @404-Page-Found with @Copilot in https://github.com/CherryHQ/cherry-studio/pull/23814
- feat(export): label assistant messages with model name in Obsidian export by @404-Page-Found with @Copilot in https://github.com/CherryHQ/cherry-studio/pull/23584
- fix(agent-composer): keep model controls available while editing by @WhiteZeek in https://github.com/CherryHQ/cherry-studio/pull/24694
- fix(chat-pane): keep the browser resize bar draggable at full height by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24701
- fix(heartbeat-editor): stop polling heartbeat status while the dialog is closed by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24680
- chore(deps): drop adm-zip in favor of archiver, node-stream-zip and jszip by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24679
- refactor(empty-state): remove the no-op preset prop by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24683
- fix(binary-manager): look up system mise without blocking the main process by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24662
- feat(provider-registry): add new Flash model by @Contentment003111 in https://github.com/CherryHQ/cherry-studio/pull/24624
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24702
- fix(file): record default-app open handoffs at info level by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24548
- fix(doctor): load only the incident’s preceding messages by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24647
- fix(e2e-regression): judge a —grep rerun by the selection its own run recorded by @cherry-qa[bot] in https://github.com/CherryHQ/cherry-studio/pull/24681
- fix(notes): restore […] and (…) math in the Notes editor by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24576
- fix(agent-session): recover a directory occupying the durable store name by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24374
- fix(usage): recover provider cache buckets and label computed costs as estimates by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24715
- fix(model-check): probe streaming-capable models with complete streams by @WhiteZeek in https://github.com/CherryHQ/cherry-studio/pull/24724
- feat(command-policy): parse shell invocations and apply JSON approval rules by @1lck in https://github.com/CherryHQ/cherry-studio/pull/24434
- feat(error-reporting): classify failures before Sentry capture by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/21720
- fix(cherry-cloud): let the server decide quota refusals instead of the client by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24687
- docs(agent-prediction): specify native runtime suggestions by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24717
- feat(agent-runtime): use native prompt prediction for Claude and Pi by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24722
- fix(empty-state): explain empty assistant, agent and note sidebars by @SiinXu in https://github.com/CherryHQ/cherry-studio/pull/24684
- fix(open-mineru): support MinerU 4.1 Tasks parsing by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24745
- feat(provider-registry): support GPT-6.2 Astra and Sol by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24749
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24752
- feat(zoom): support Ctrl + mouse wheel to zoom the interface by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24730
- docs(file-preview): specify file link line navigation by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24746
- chore(lint-tooling): adopt tsrs for lint and type checks by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24747
- feat(file-preview): navigate file links to source lines by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24750
- fix(window-manager): clear per-window state when pooled windows return to the pool by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24162
- fix(changeset-policy): reject ineligible and unchanged packages by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24773
- fix(file-observation): keep file trees and previews consistent by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24728
- fix(model-parameters): enforce active model sampling limits by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24776
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24785
- fix(native-rebuild): drop the obsolete rebuild:node step by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24789
- chore(lint): enforce import layout with oxlint by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24790
- feat(provider-registry): expose Mistral Large 4 reasoning controls by @404-Page-Found with @Copilot in https://github.com/CherryHQ/cherry-studio/pull/23236
- feat(file-preview): PPTD slide preview plugin by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24796
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24800
- fix(agent-proxy): keep unparseable IPv6 entries out of the Agent NO_PROXY by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24793
- feat(model-speed-control): explain service tiers and show saved vs effective selection by @404-Page-Found in https://github.com/CherryHQ/cherry-studio/pull/24788
- chore(deps): drop redundant @types/uuid by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24810
- fix(knowledge): validate source URLs with the shared http(s) check by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24812
- fix(selection-agent): stop a stale session from resizing a reused window by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24811
- feat(knowledge): accept multiple URLs in one add-dialog submission by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24794
- feat(paintings): show the model and returned size of a generated image, and save the original file by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24807
- fix(api-gateway): report overflow token counts in Claude Code’s format by @kovsu in https://github.com/CherryHQ/cherry-studio/pull/24830
- chore: version packages by @github-actions[bot] in https://github.com/CherryHQ/cherry-studio/pull/24832
- fix(agent-session): bound trace and native observation memory by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24500
- docs(agent-notes): propose agent image byte paths spec by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24650
- perf(agent-runtime): share frozen observation values instead of copying history by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24651
- perf(file-processor): validate each native history image once by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24661
- perf(agent-session): release finished Durable subagents by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24705
- perf(agent-runtime-pi-durable): read only new Durable history on repeated reads by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24714
- docs(agent-notes): propose deriving Durable facts and observations as folds by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24813
- perf(agent-runtime-pi-durable): derive behavior facts from in-memory folds by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24814
- perf(agent-runtime-pi-durable): publish observation deltas from a history fold by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24817
- perf(agent-session): apply Durable deltas without per-update rescans by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24818
- docs(pi-durable): propose Durable turn lifecycle authority by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24759
- fix(pi-durable): report a Durable abort as a paused turn by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24760
- fix(pi-durable): keep the Durable access open on Stop by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24766
- fix(pi-durable): recover interrupted replies from the Durable record by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24774
- fix(pi-durable): admit busy-time input in the Durable queue by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24777
- test(pi-durable): verify turn authority in Electron and mark the Spec implemented by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24781
- docs(agent-import): specify DeepSeek import into Agent sessions by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24729
- refactor(agent-runtime): replace internal HTTP codecs with protocol objects by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24786
- test(agent-session): fix historical tree migration fixtures by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24851
- feat(agent-import): import DeepSeek history into Agent sessions by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24738
- docs(conversation-list): specify independent sorting by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24836
- feat(conversation-list): separate grouping and sorting by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24838
- fix(topics): keep restored conversations openable from the sidebar by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24852
- fix(chat): keep nested scrollers hittable while the message list scrolls by @cherry-coding[bot] in https://github.com/CherryHQ/cherry-studio/pull/24848
- fix(release): backport #24853 to v2.2.3 by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24855
- fix(release): backport #24859 to v2.2.3 by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24860
- fix(release): backport #24876 to v2.2.3 by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24877
- test(pi-durable): backport overflow budget to release v2.2.3 by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24880
- fix(agent-session): backport #24913 to v2.2.3 by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24935
- fix(release): backport #24919 to v2.2.3 by @kangfenmao in https://github.com/CherryHQ/cherry-studio/pull/24941
- fix(cherry-cloud): backport #24731 to v2.2.3 by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24943
- test(pi-durable): backport #24945 to v2.2.3 by @DeJeune in https://github.com/CherryHQ/cherry-studio/pull/24946
New Contributors
- @markiceman made their first contribution in https://github.com/CherryHQ/cherry-studio/pull/24494
- @zuhairkhalid229 made their first contribution in https://github.com/CherryHQ/cherry-studio/pull/24603
- @JeremySu0818 made their first contribution in https://github.com/CherryHQ/cherry-studio/pull/24595
Full Changelog: https://github.com/CherryHQ/cherry-studio/compare/v2.2.2…v2.2.3
Source CherryHQ/cherry-studio ·
v2.2.3· 2026-11-19 · upstream release
Mirrored verbatim for indexing. All rights belong to the upstream project.
Release notes are published by the upstream project and mirrored here for indexing. All rights belong to their authors.