What's Changed
- enterprise docs by @akshaydeo in #6298
- feat(logstore): add
DimensionScopeceiling to bound grouping dimension values per caller by @impoiler in #6262 - refactor: extract routing rules and complexity router into a dedicated routing plugin by @Pratham-Mishra04 in #6144
- docs: add
claimsSyncModeprovisioning source section to all SCIM provider guides by @BearTS in #6294 - refactor: move
PublishRoutingAllowlistandLoadBalanceProviderfrom governance'sPreRequestHookinto the routing plugin so both run after rule evaluation on the post-rule model by @Pratham-Mishra04 in #6145 - refactor: extract routing rules and complexity analyzer endpoints into dedicated
RoutingHandlerunder/api/routing/*with backwards-compatible/api/governance/*aliases by @Pratham-Mishra04 in #6146 - docs: migrate routing rules and complexity analyzer endpoints from
/api/governance/*to/api/routing/*with deprecated legacy aliases by @Pratham-Mishra04 in #6147 - docs: bedrock application inference profiles by @TejasGhatte in #6203
- fix: default http2 timeout interval in ui by @TejasGhatte in #6228
- dependabot fixes by @akshaydeo in #6305
- deendabot fixes by @akshaydeo in #6267
- go upgrade by @akshaydeo in #6269
- feat: add fallback UI and logos for Splunk connector by @roroghost17 in #6091
- docs: Add Splunk connector docs by @roroghost17 in #6099
- feat: add config json and helm charts support for Splunk by @roroghost17 in #6296
- feat: datasheet-backed compatibility flows by @TejasGhatte in #6281
- fix: bedrock tool use ids by @TejasGhatte in #6300
- fix: rerank integration by @TejasGhatte in #6301
- tests: harness for runware by @TejasGhatte in #6303
- refactor: updates modelcatalog to memoise provider configs recomputing for allowed models / providers by @roroghost17 in #5641
- refactor: fixes some minor unnecessary allocations majorly in core by @roroghost17 in #5657
- perf: rewrites individual span writes to bulk and reuses delivery timer by @roroghost17 in #5956
- perf: alloc reductions, span attribute map reuse, and logging context read on final chunk by @roroghost17 in #5957
- feat: add Bifrost overhead latency to logs details view by @roroghost17 in #5533
- feat: adds overhead data to matviews and db by @roroghost17 in #5534
- feat: adds bifrost overhead chart to UI dashboards by @roroghost17 in #5535
- feat: adds status code tags to LLM error logs by @roroghost17 in #5536
- fix: adds plugin logs existence check before draining by @roroghost17 in #6211
- fix: resolve merge conflicts in migration tests, version, and logs page by @impoiler in #6318
- refactor: add gencache package and wire in GetModelsForProvider to gencache by @roroghost17 in #6224
- perf: optimize ordered map's marshalJson by @roroghost17 in #6241
- perf: cache serialized json for shared mcp tools by @roroghost17 in #6242
- perf: update json encode usage with sonic marshal by @roroghost17 in #6268
- compilation issues by @akshaydeo in #6321
- onbaording checklist fixes by @akshaydeo in #6322
- Preserve Azure Responses stream errors after HTTP 200 by @dani29 in #6302
- feat: add megapixel-tier image pricing fields (4/8/16/32/64MP) across DB, cost engine, overrides, API, docs, and UI by @Pratham-Mishra04 in #6082
- feat: backfill upscale output resolution from
target/factorparams andmetrics.resolution_targetfor accurate resolution-tiered pricing by @Pratham-Mishra04 in #6083 - fix: use
strings.Containsfor "together" provider normalization to handle substrings by @Pratham-Mishra04 in #6320 - tests: fixes runware harness tests by @TejasGhatte in #6319
- fix(ui): move beta badge into
PageTitle, simplify notification trigger logic, and fix virtual keys toolbar overflow by @impoiler in #6324 - fix: adds gh workflow to spin up vllm instances for testing vllm by @sammaji in #5541
- fix: adds openrouter stt / tts by @sammaji in #5734
- fix: include openrouter embedding models in list models call by @sammaji in #6264
- fix: register bedrock mantle in ProviderSendsDoneMarker by @sammaji in #6021
- fix: route grok models through openai/v1 path for bedrock mantle by @sammaji in #6022
- fix: force reasoning effort to none when reasoning with tools is unsupported by @sammaji in #6293
- fix(ui): truncate long values in log detail view and improve provider button responsiveness by @impoiler in #6330
- docs(migration): consolidate governance API namespace guide into v2.0.0 migration page by @impoiler in #6332
- feat: add batch schemas and provider plumbing by @SahilChoudhary22 in #5291
- feat: add batch pricing to model catalog by @SahilChoudhary22 in #5292
- feat: add batch jobs table to configstore and batch debug to logstore by @SahilChoudhary22 in #5293
- autodetect harness level session ids by @akshaydeo in #6333
- feat: add batch accounting engine and sweeper by @SahilChoudhary22 in #5294
- feat: report batch usage to governance by @SahilChoudhary22 in #5295
- feat: wire batch accounting into logging and transport by @SahilChoudhary22 in #5296
- fix: handle log cost recalculation for batch by @sammaji in #6121
- feat: adds row in logs sheet to render batch pricing information by @sammaji in #6109
- chore: adds missing batch / batch pricing tests in harness by @sammaji in #6299
- tests: anthropic message start harness test by @TejasGhatte in #6327
- fix: rerank compatibility and integration tests by @TejasGhatte in #6328
- fix: close leaked postgres pools and stale test date in logstore tests by @sammaji in #6351
- fix: restructuring of replayed thinking blocks by @akshaydeo in #6346
- gemini inmessage system role bug fix by @akshaydeo in #6363
- fix(ui): unify full-height page sizing with
--app-bottom-paddingCSS variable and tighten sidebar/content margins by @impoiler in #6370 - fix: handle dashboard auth, budget-reset dialog, and stale MCP state in e2e tests by @sammaji in #6364
- feat: adds input output cost split for inference calls by @roroghost17 in #4856
- feat: adds denormalized cost columns for logs by @roroghost17 in #4857
- feat: add BifrostCost support to speech, transcription and ocr usages by @roroghost17 in #6338
- fix: populate cost breakdown on fallback billing paths by @roroghost17 in #6339
- feat: expose input/output/additional cost split in quota API by @roroghost17 in #6340
- fix: adds split cost calculation to Clickhouse and recalculation flows by @roroghost17 in #6343
- feat: wire in semantic cache cost to additional cost by @roroghost17 in #6359
- fix: rerank costs and integration tests for langchain by @TejasGhatte in #6358
- fix: runware inputs and docs for operations by @TejasGhatte in #6366
- fix: adds batch types in filters and make batch id column take full width by @sammaji in #6376
- fix: redact identity-aware-proxy headers before connector export by @roroghost17 in #6371
- feat: filter
Accept-Encodingto supported codecs on passthrough headers and expandCheckAndDecodeBodyto handle deflate, brotli, zstd, and chained encodings by @Pratham-Mishra04 in #6360 - docs: add v1.5.x downgrade guide for deployments that ran
v2.0.0-prerelease3withent_split_oidc_session_auth_token_columnmigration by @Pratham-Mishra04 in #6374 - feat: add
HTTPTransportPreAuthHookphase that runs before transport authentication, movingHTTPTransportPreHookto run after by @Pratham-Mishra04 in #6375 - feat: add
virtual-key-from-confignative Go plugin example with pre-auth hook, tests, and README by @Pratham-Mishra04 in #6373 - feat: runware list models api by @TejasGhatte in #6372
- feat: add per-plugin
semaphore_sizeandinject_timeouttoPluginConfigwith context-boundedInjectcalls and tracer-default fallbacks by @BearTS in #6341 - fix: build fix for scimAPI fallback by @roroghost17 in #6384
- chore: update flake for 1.26.7 by @roroghost17 in #6385
- fix: bedrock langchain embedding by @TejasGhatte in #6381
- feat: start sending overhead metrics to connectors by @roroghost17 in #6345
- perf: carry the *Span pointer on span handles by @roroghost17 in #6387
- feat: per-phase overhead breakdown - instrument, compute, persist, display by @roroghost17 in #6388
- feat: keep breakdown spans out of observability connectors by @roroghost17 in #6389
- [Fix]: Redis vectorstore - escape all RedisSearch special characters in TAG query values by @AdityaPainuli in #5351
- feat: add per-size and joint size+quality image pricing columns, cost computation, and overrides for
gpt-image-1-style models by @Pratham-Mishra04 in #6379 - feat: add per-size and joint size+quality image pricing fields for 1024×1536 and 1536×1024 resolutions by @Pratham-Mishra04 in #6380
- chore: add
openai_configref to all transport provider config schemas by @BearTS in #6275 - fix: openai service tier ultrafast by @TejasGhatte in #6396
- fix: update overrides fields by @TejasGhatte in #6399
- chore: sync v1.6.x to dev by @impoiler in #6401
- tests: harness tests fixes by @TejasGhatte in #6408
- refactor: extract shared MCP client form fields, validation, and payload builder into
mcpClientFormFieldsand reuse across create and library install sheets by @Pratham-Mishra04 in #6411 - feat: add broker clustering support to Helm chart by @danpiths in #6398
- fix: accept json bodies in image edit request by @TejasGhatte in #6418
- fix: make tool_sync_interval to be in minutes by @sammaji in #6409
- fix: logs in batches ui, resolve VK-scoped wildcard budgets when request has no model
by @sammaji in #6410 - chore: remove legacy metrics in connectors by @roroghost17 in #6403
- fix: update stream overhead with sub-buckets and UI fixes by @roroghost17 in #6433
- chore: fix indentation and replace
overflow-x-hiddenwithoverflow-hidden!inmcpClientSheetby @Pratham-Mishra04 in #6428 - feat: project
needs_reauthonto shared-OAuth clients when their token row is invalidated via batchGetSharedOauthTokensByConfigIDslookup by @Pratham-Mishra04 in #6429 - feat: pre-flight verify and live-reconnect sticky MCP HTTP client on static header update, with
mcpHeadersEqualdiff guard and tests by @Pratham-Mishra04 in #6430 - fix: park failed
EnableClientatDisabledinstead ofUnstableto prevent retry wedging, addErrMCPEnableConnectFailed, and guardisEnableableon both state and config by @Pratham-Mishra04 in #6431 - docs: require cluster mode for multi-replica, document Consul auth/TLS and lifecycle by @R-droid101 in #6415
- fix: resolve Together and alias pricing in management catalog by @dani29 in #6257
- tests: integration for rerank langchain by @TejasGhatte in #6432
- docs: fix Discord badge rendering on Glama by @Swpn0neel in #6425
- fix: preserve provider key name on update, keep constraint detail in already-exists errors by @cpsc in #6417
- fix: use Glama-compatible Discord badge by @Swpn0neel in #6448
- chore: update core with sub-buckets by @roroghost17 in #6470
- move version into the dropdown by @akshaydeo in #6476
- add support for skip provider checks by @akshaydeo in #6479
- fix: use bifrost as user-agent name, show pricing for manual batch calls when resolved by sweeper by @sammaji in #6474
- [fix]: HuggingFace provider - default stream_options.include_usage on chat streaming by @elliottrabac in #6478
- [fix]: core/providers/openai - preserve max_tokens for OpenCode endpoints by @Alex-wangyang in #6458
- docs: add v2.0.0 availability callouts across features, providers, and API references by @impoiler in #6420
- fix(ui): add open-state highlight to topbar triggers, inline
sideOffset, and removeTOPBAR_MENU_SIDE_OFFSETconstant by @impoiler in #6485 - fix: refresh tools synchronously on per-call shared-credential client credential update instead of returning
ErrMCPReconnectNotApplicable, with disabled-client guard and per-user sentinel preserved by @Pratham-Mishra04 in #6483 - fix: replace
maps.Copymerge with wholesale replacement inSetClientToolsandUpdateClientCredentialsto prevent removed tools from persisting in memory by @Pratham-Mishra04 in #6484 - fix: vertex batches anthropic compatibility by @TejasGhatte in #5368
- fix: model caps check for effort by @TejasGhatte in #6492
- topbar fixes by @akshaydeo in #6493
- fix: update server tool calls in log details ui by @TejasGhatte in #6475
- fix: remove "disable sync" semantic from
tool_sync_interval(negative now rejected), addUpdateMCPToolSyncIntervalfor hot-reload, restart per-call checker onEnableClient/sticky→per-call flip, and re-time running checkers on global or per-client in by @Pratham-Mishra04 in #6502 - fix: unify
getAuthScopeDisplaywithauthScopeOffrom form fields, usingMCPAuthTypeinstead ofstringby @Pratham-Mishra04 in #6504 - fix: stop background cost recalculation zeroing batch aggregate rows and charge access profile budgets for model-less requests by @sammaji in #6505
- chore: fix streaming time calculations by @roroghost17 in #6495
- feat: update log details views UI to show input / output cost split by @roroghost17 in #6511
- docs: updated migration guide for 2.0.0 with the cost split function by @roroghost17 in #6515
- fix: OTEL UI fix by @roroghost17 in #6518
- [fix]: core/providers/opencode - route Responses requests to /v1/resp… by @mohammadrezwankhan in #6503
- feat(helm): add nodeSelector/tolerations/affinity to hosted PostgreSQL by @CarlosLanderas in #6506
- fix: exclude batch echo rows from missing-cost filter, prevent recalculation billing them, and enforce model allowlist for inline batch requests by @sammaji in #6522
- fix: replace fixed 500ms sleep with 5s polling loop in Redis vector search test to avoid flakiness under parallel test runs by @Pratham-Mishra04 in #6524
- chore: release v2.0.0 by @Pratham-Mishra04 in #6526
- chore: add "One Entry Per Feature" rule to changelog-writer skill, folding in-window follow-up fixes into feature entries and reserving
fix:for pre-existing bugs by @Pratham-Mishra04 in #6527 - fix: drop service_tier for Bedrock Mantle instead of forwarding a field it rejects by @zachgersh in #6532
- chore(ui): remove Evals sidebar item, fix template literal formatting, and reindent nested arrays by @impoiler in #6531
- go 1.27.0 upgrade by @akshaydeo in #6535
- feat(helm): support for external PostgreSQL support for logs store in helm by @BearTS in #6226
- chore: condense
values.yamlcomments for clarity and brevity by @BearTS in #6509 - chore: remove
deploymentsfield from Azure/Vertex schema and replace withmodels/aliasesin values.yaml comments by @BearTS in #6510 - chore: restructure guardrails
values.yamlto listprovidersbeforeruleswith expanded inline examples andprovider_config_idslinkage by @BearTS in #6512 - chore: add
metrics_enabledboolean to Bifrost Helm chart metrics schema and values by @BearTS in #6513 - chore: inherit semantic cache embedding keys from
bifrost.providersand dropkeysfield from semantic cache config across Helm charts, values, and docs by @BearTS in #6517 - fix: skip key selection on anthropic provider by @TejasGhatte in #6536
- fix: clear BifrostContextKeyURLPath and BifrostContextKeyDirectKey by @TejasGhatte in #6538
- chore: release Bifrost Helm chart v2.1.37 with Splunk HEC plugin, OTel/logging semaphore controls, PostgreSQL node scheduling, and separate logs-store Postgres by @BearTS in #6542
- routing plugin by @akshaydeo in #6545
- fix go version by @akshaydeo in #6546
- patched missing file embedding by @akshaydeo in #6544
- 2.0.0 by @akshaydeo in #6547
- fix: OSS build fix by @akshaydeo in #6551
- chore: remove
deploymentsfield from Azure/Bedrock config schemas and validation scripts by @BearTS in #6554 - 2.0.0 test fixes by @akshaydeo in #6552
- chore: remove
keysfield from semantic cache Helm config validation script by @BearTS in #6555 - release pipeline updates by @akshaydeo in #6556
- docs: langchain rerank and base url updates by @TejasGhatte in #6558
- fix: fixes openai integrations cost usage by @roroghost17 in #6557
- core version bump by @akshaydeo in #6560
- Test fixes by @akshaydeo in #6561
- changelog patches by @akshaydeo in #6563
- changelog patches (#6563) by @akshaydeo in #6564
- sonic update, test case fixes by @akshaydeo in #6565
- aug-26-test-case fixes 2 by @akshaydeo in #6566
- core test fixes by @akshaydeo in #6568
- fix: update new attribute names in OTEl by @roroghost17 in #6567
- fixes API tests by @akshaydeo in #6569
- fix: Anthropic Compaction Token Undercounting, Resulting In Big Output Token Count Miss by @17jmumford in #6500
- adds core version bump by @akshaydeo in #6572
- anthropic server side fallback token compute fix by @akshaydeo in #6573
- routing plugin go mod fixes by @akshaydeo in #6581
- merge commit by @akshaydeo in #6584
- Main backmerge by @akshaydeo in #6585
- routing plugin go mod fixes (#6581) by @akshaydeo in #6582
New Contributors
- @Swpn0neel made their first contribution in #6425
- @cpsc made their first contribution in #6417
- @elliottrabac made their first contribution in #6478
- @Alex-wangyang made their first contribution in #6458
- @mohammadrezwankhan made their first contribution in #6503
- @CarlosLanderas made their first contribution in #6506
Full Changelog: ent-v1.5.12-base...ent-v2.0.0-base