onyx-dot-app/onyx v2.7.0

on GitHub

See the assets to download this version and install.

What's Changed

• refactor: let litellm handle translating reasoning_effort to anthropic thinking by @wenxi-onyx in #7090
• feat(ux): include a copy button for chat stack traces by @jmelahman in #7091
• fix: think tool newline unescaping by @rohoswagger in #7086
• refactor: llm provider forms by @Weves in #7006
• fix: get_tenant_users script invalid sql stmt by @wenxi-onyx in #7097
• chore: Remove unused resources from model server by @yuhongsun96 in #7094
• fix: The update method for the new Vespa interface should correctly handle None chunk_count by @acaprau in #7098
• chore(fe): make Text component default to span by @jmelahman in #7096
• chore: docker compose no MCP server by @yuhongsun96 in #7100
• chore(fe): Text default span follow up by @jmelahman in #7106
• chore: Fix Tests by @yuhongsun96 in #7107
• chore(fe): refresh chat Stack Trace button by @jmelahman in #7092
• chore(docker): add healthchecks by @jmelahman in #7089
• chore: Add opensearch-py 3.0.0 by @acaprau in #7103
• chore(desktop): remove duplicate startup log, onyx-desktop by @jmelahman in #7113
• chore(desktop): disable nightly builds by @jmelahman in #7115
• refactor: Improve form field components by @raunakab in #7104
• refactor(ux): Share Chat modal uses CopyIconButton by @jmelahman in #7116
• chore(fe): fix some Text that should be spans by @jmelahman in #7112
• perf: update web/STANDARDS.md + add standards to CLAUDE.md / AGENTS.md by @Weves in #7039
• feat: add z indices for confirmation modal by @wenxi-onyx in #7114
• fix: stop-generation for non-deep research by @rohoswagger in #7045
• fix: rendering parallel research agents cleanly by @rohoswagger in #7078
• fix: popover max height setting by @rohoswagger in #7093
• chore(gha): ensure uv cache is pruned before upload by @jmelahman in #7120
• fix: MIT integration tests by @yuhongsun96 in #7121
• chore: clean up stop signal redis fence by @rohoswagger in #7119
• fix: stop-generation for deep research by @rohoswagger in #7050
• chore: Add script to restart OpenSearch container by @acaprau in #7110
• feat: Add OpenSearch schema by @acaprau in #7118
• chore: prompts by @yuhongsun96 in #7108
• fix: gemini default location global by @wenxi-onyx in #7124
• fix: persist onboarding flow until user explicitly finishes by @devin-ai-integration[bot] in #7111
• chore(fe): baseline align inline code spans by @jmelahman in #7128
• fix: suppress Jest act() warning spam in test output by @nmgarza5 in #7127
• fix: reenable force search parameter by @yuhongsun96 in #7130
• refactor: onboarding forms by @Weves in #7105
• fix: Remove Default Reminder by @yuhongsun96 in #7131
• fix: downgrade by @Weves in #7132
• fix: lazy load tracing providers to avoid spamming logs when not configured by @wenxi-onyx in #7134
• fix: make litellm testing script prettier by @wenxi-onyx in #7136
• fix: openai provider identification on the admin panel by @Weves in #7135
• feat: Refreshed agent creation page by @raunakab in #6241
• Fix/input combobox dropdown by @Subash-Mohan in #7015
• refactor(modals): All modals use new Modal component by @Subash-Mohan in #6729
• Feat/image config backend by @Subash-Mohan in #6961
• Update README.md by @yuhongsun96 in #7140
• Chat Flow Readme by @yuhongsun96 in #7142
• fix: prevent Slack federated search query multiplication by @nmgarza5 in #7125
• fix(desktop): Disable reload on Mac by @justin-tahara in #7141
• feat: add auto LLM model updates from GitHub config by @Weves in #6830
• feat(connectors): Add Deletion Popup by @justin-tahara in #7054
• fix: skip failing tests by @Weves in #7152
• chore: fix linter issues by @yuhongsun96 in #7122
• fix: onboarding display by @Weves in #7153
• feat(tools): extend open_url to handle indexed content urls by @jessicasingh7 in #6822
• feat: add new models to cloud by @Weves in #7149
• fix: when onboarding flow shows up by @Weves in #7154
• fix: Duplicate model provider sections for unenriched LLM models by @nmgarza5 in #7148
• refactor: White-labelling by @Subash-Mohan in #6938
• fix: check stop signal during active streaming by @devin-ai-integration[bot] in #7151
• feat: add tool call validation to eval cli by @rohoswagger in #7144
• feat: running evals locally by @rohoswagger in #7145
• perf(chat): consildate chat UI layout style by @jmelahman in #7129
• fix: non-thinking responses not displaying until page refresh by @devin-ai-integration[bot] in #7123
• perf(chat): memoize chat messages by @jmelahman in #7157
• chore(fe): follow up fixes to #7129 by @jmelahman in #7160
• chore(deps): bump qs from 6.14.0 to 6.14.1 in /web by @dependabot[bot] in #7147
• refactor(whitespace): rm temporary react fragments by @jmelahman in #7161
• chore(blame): introduce .git-blame-ignore-revs to ignore refactors by @jmelahman in #7162
• Tests/theme by @Subash-Mohan in #7163
• feat(config): Image generation frontend by @Subash-Mohan in #7019
• feat(image-generation): e2e tests by @Subash-Mohan in #7164
• chore(docker): publish inference_model_server port 9000 in dev by @jmelahman in #7166
• fix: don't pass selected tab to connector specific config by @wenxi-onyx in #7165
• fix(ollama): rm unsupported tool_choice option by @jmelahman in #7156
• feat(opensearch): Add OpenSearch queries by @acaprau in #7133
• fix(braintrust): Updating naming for metric by @justin-tahara in #7168
• chore: cleanup old LLM provider update mechanism by @Weves in #7170
• chore(docker): .dockerignore /tests/ by @jmelahman in #7172
• fix(braintrust): Fix Tenant ID to Token Association by @justin-tahara in #7173
• chore(playwright): prefer baseURL by @jmelahman in #7171
• chore(deployments): prefer release environment by @jmelahman in #6997
• feat(opensearch): Add OpenSearch client by @acaprau in #7137
• chore(opensearch): Add external dep tests for OpenSearchClient by @acaprau in #7155
• chore(playwright): fix Exa configure tests by @jmelahman in #7176
• fix(braintrust): Span Attributes Association by @justin-tahara in #7174
• New send message api by @yuhongsun96 in #7167
• chore: Plugin issue by @yuhongsun96 in #7179
• revert: "chore(pre-commit): run uv-sync in active venv" by @jmelahman in #7178
• chore: spacing by @yuhongsun96 in #7183
• fix: regenerate models stuck in perma loading state by @Weves in #7182
• feat: nonstreaming send chat message api by @yuhongsun96 in #7181
• feat: improve display for models that are no longer present by @Weves in #7184
• feat: Easy send message nonstreaming by @yuhongsun96 in #7186
• chore(fe): improve human chat responsiveness by @jmelahman in #7187
• chore(fe): rm unnecessary spacer from chat ui by @jmelahman in #7189
• refactor(whitespace): rm react fragment by @jmelahman in #7190
• chore(whitespace): ignore refactor rev by @jmelahman in #7191
• feat: Deep Research Internal Search Tuning by @yuhongsun96 in #7193
• feat: cloud usage limits by @evan-onyx in #7192

Full Changelog: v2.6.0...v2.7.0