upsert_deployment, add_deployment, delete_deployment and update_settings run on the Python side that normalizes deployments, then swap the native snapshot; running calls keep theirs. A change the Rust backend cannot serve hands the facade over to a PythonRouter for good, or raises under RUST_REQUIRED
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
A streamed Anthropic message holds its lifecycle frames until its first content delta, deciding with PythonRouter's own gates whether a retry or fallback could take over. An error frame or refusal before then retries in the same group with Python's budget rules, then falls back on the hop's chain. Retry decisions and cooldowns judge the provider error inside the mid-stream envelope
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
Both go through a mirror of the generic helper after selection, with metadata in litellm_metadata. A Responses stream holds its opening lifecycle events until its first output and falls back from a drop before then on its hop's chain. A drop after output resumes the chain through a new Engine::resume entry point, continuing from the generated text. Python's Responses stream wrapper class moves to router_utils so both backends share it. Streaming Anthropic messages still raises NotImplementedError
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
A stream failing before content now settles the call at the hop it failed in, as Python's stream wrappers do: the outer retry and fallback layers no longer retry it, try their next target or rewrite its outcome, and the caller gets the provider error rather than the MidStreamFallbackError. The sync driver buffers and replays streams too and follows the async rule, since Python's sync wrapper reruns the same group and can recurse until RecursionError. The sync path no longer updates the call counters, matching Python
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
The call driver reads an attempt's stream until its first content chunk and hands back a
ReplayStream that replays what it buffered. A MidStreamFallbackError before content goes
straight to the fallback chain without retrying the group, as Python's chat stream wrapper
does; any other error before content is raised as is, and after content a mid-stream
fallback error surfaces as its original exception.
A native test runs the same failing-before-first-chunk fallback chain through both backends
and gets the same content from the same sequence of model groups.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
Each scenario runs the same config and failures through PythonRouter and the Rust backend
against one redis-server and compares every key, raw stored value (write time masked) and
TTL: deployment cooldowns, allowed_fails counters and failure RPM keys. A Rust router also
honors cooldowns a Python router wrote. Skipped when redis-server is not installed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
pinned_backend fixes the decision Router(...) makes, and the router_backend fixture runs an
opted-in module once per backend, with the Rust run pinned to RUST_REQUIRED so a config it
cannot serve fails instead of quietly running on Python. The per-deployment num_retries and
exception redaction suites are opted in; the Rust-side failures left are listed in RUST_GAPS
as strict xfails (tests of PythonRouter internals, image generation, dict fallback targets).
Closing gaps the run surfaced: mock_testing_rate_limit_error replaces the first attempt of
the first hop, mock_timeout reaches first-hop attempts only, the call counters are readable,
and fallback keys containing a slash decline only when Python would match them to a group by
provider prefix.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
test_aadapter_completion patched _aadapter_completion on the facade, which the backend never
calls, and test_init_auto_router_deployment_success expected the facade where the backend
passes itself to AutoRouter. Both now target PythonRouter, as the facade refactor did for
the other tests.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
python-bridge exports Router, which holds the routing engine and runs each call through
host-python's run_call. Every attempt is an Invoke host op and every backoff a Sleep host
op; the async driver awaits them inline in the caller's task and the sync driver answers
them directly.
RustRouter normalizes deployments with a callback-free PythonRouter built from the same
arguments, projects deployments and resolved settings to Rust, and answers each Invoke with
RoutedCall, which runs _acompletion's post-selection body, classifies what litellm raised
and applies the router's metadata and exception ops. acompletion and completion are served;
the request options and members it does not cover yet raise NotImplementedError naming them.
Configs with provider-prefixed or dict fallback entries, or callbacks that observe fallback
events, now decline to PythonRouter.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
Router(...) now resolves Route.ROUTER once at construction. PYTHON builds PythonRouter as
before. Under RUST_OPT_IN (LITELLM_RUST) the Rust backend is built when it can serve the
config and the call declines to PythonRouter otherwise; RUST_REQUIRED raises naming the
unsupported argument, deployment field or global setting.
rust_support decides what the first increment can serve: the reliability arguments
(retries, fallbacks, retry policies, cooldowns, Redis) on simple-shuffle, without
wildcards, strategy routers, orders, tags, tpm/rpm limits, team deployments, aliases or
callbacks that filter deployments. RustRouter is a stub until the native router exists, so
every config still declines today.
Also types the facade's conditional base as a TypeAlias and marks its forwarding kwargs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Aw1xktpUnix5tWhvSzey8t
The backend passes itself, not the facade, into helpers like cache-aware routing, so the
`proxy_server.llm_router is router` check never matched and cache-aware routing turned
itself off. Tests asserting router identity or reading backend-only state now go through
`router.backend`.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VBPzD1X9jBTmKn6ypDQQTY
First step of the Rust router POC. The existing implementation moves unchanged to
litellm/router_backends/python_router.py as PythonRouter, and litellm.Router becomes a
facade that holds it as `backend` and forwards the full current surface, so a Rust
backend can sit behind the same class later. No behavior change.
Tests that override or patch hooks the backend calls on itself now target PythonRouter,
and imports of private helpers and module-global patch strings point at the new module.
litellm-rust/crates/router/POC.md is a hand-off copy of the POC design for sessions that
cannot see the local notes. Delete it before opening a PR.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VBPzD1X9jBTmKn6ypDQQTY
* feat(lens): add dataset case and size limits
* feat(lens): add dataset, case and build models
* feat(lens): build dataset cases from traces, findings and text
* feat(lens): store dataset revisions insert-only
* feat(lens): add dataset routes for build, save, export and eval cases
* feat(lens): mount the dataset router before lens routes
* feat(lens): add LiteLLM_LensDataset table
* feat(lens): add LiteLLM_LensDataset table to proxy schema
* feat(lens): add LiteLLM_LensDataset table to extras schema
* feat(lens): add migration that creates the dataset table
* test(lens): cover dataset case building, dedupe and limits
* test(lens): cover dataset revisions, conflicts and eval cases
* chore(ui): regenerate API types for lens datasets
* feat(lens): add dataset UI types
* feat(lens): add datasets API client
* feat(lens): add dataset query and mutation hooks
* feat(lens): add case selection and expected edit logic
* test(lens): cover case selection and expected edits
* feat(lens): add the add to dataset dialog
* test(lens): cover saving picked cases from the dialog
* feat(lens): add datasets list
* feat(lens): add dataset detail with revisions and export
* test(lens): cover editing, revisions and export in datasets tab
* feat(lens): expose datasets on the lens API
* feat(lens): add in-memory datasets for demo mode
* feat(lens): wire demo datasets into the demo lens API
* feat(lens): add optional lens API hook
* feat(lens): add optional onboarding hook
* feat(lens): add datasets tab and dataset routing
* feat(lens): show the datasets tab
* feat(lens): add to dataset from the trace header
* feat(lens): add a single turn to a dataset from a step
* feat(lens): add finding evidence to a dataset
* docs(lens): add datasets screenshots for the PR
* test(lens): cover dataset revision storage against Postgres
* test(lens): cover dataset trace paging, findings and route errors
* test(lens): cover dataset build fallbacks and no_content skips
* fix(lens): register the dataset table for postgres span names
* feat(lens): add invalid skip reason for unparseable case lines
* fix(lens): keep valid JSONL cases, provenance and size limits; stop reading past the case cap
* test(lens): cover malformed JSONL, re-import provenance, size fields and early cap
* chore(ui): regenerate API types for the invalid skip reason
* feat(lens): show text for the invalid skip reason
* feat(lens): render datasets in the same inspector table as investigations
* test(lens): open a dataset by clicking its table row
* docs(lens): update the datasets list screenshot
* style(lens): format the datasets table
* fix(lens): normalize JSON text span input and output into messages when building cases
* test(lens): cover JSON text span normalization for dataset cases
* Revert "fix(lens): normalize JSON text span input and output into messages when building cases"
This reverts commit 29c3e34962fa12427dbd516c08f2599db997a905.
* fix(traces): normalize agent assistant summaries into UI messages
* feat(lens): add dataset case view helpers built on the trace parsers
* test(lens): cover dataset case view helpers
* feat(lens): show dataset cases in an inspector table
* feat(lens): open a dataset case in a side panel with trace message cards
* feat(lens): rebuild the dataset page header and layout
* feat(lens): keep the open dataset case in the URL
* test(lens): drive dataset edits through the case table and panel
* docs(lens): update dataset view screenshots
* Revert "test(lens): cover JSON text span normalization for dataset cases"
This reverts commit 30229ae8ae90265c09d755c43c1f1eeaa66050c7.
* fix(lens): import StateMessage from shared in dataset detail
* fix(lens): import StateMessage from shared in datasets list
* fix(lens): give the datasets migration a unique timestamp after review checkpoints
* fix(proxy): resolve oidc/ pass-through credentials on every request
A pass-through credential (a use_in_pass_through deployment's api_key or
the provider's env var, e.g. TYPESAFE_API_KEY) was sent as a literal
string, so an `oidc/...` reference such as
`oidc/file//var/run/secrets/<name>/token` ended up on the wire as
`Bearer oidc/file/...`, and a rotating projected Kubernetes service
account token could not be used for pass-through auth.
Resolve credentials that start with `oidc/` through get_secret_str() on
every get_credentials() call, so the file is re-read and rotations apply
without a restart. oidc/file/ keeps its credential-directory allowlist;
other credentials are returned unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(proxy): only resolve local oidc/ pass-through credentials inline
get_credentials() runs synchronously inside async pass-through routes, so
resolving oidc/google/, oidc/github/ etc. on a cache miss would make a
blocking HTTP call to the identity provider (timeout up to 600s) on the
event loop. Limit per-request resolution to oidc/file/, oidc/env/ and
oidc/env_path/, which need no network I/O; network-backed references are
left unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(proxy): surface unreadable oidc/ pass-through credentials as 401
A missing or disallowed token file raised FileNotFoundError or
OidcPathNotAllowedError out of the route, which the proxy turned into a
generic 500 on every request. The resolver now raises an HTTPException
401 that names the credential reference and the reason, matching how the
pass-through routes already report an unset key.
oidc/env_path/ is no longer resolved inline because it reads any file
named by an environment variable without the oidc/file/ credential
directory allowlist; it is forwarded unchanged like the network-backed
references.
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: ryan-crabbe-berri <ryan@berri.ai>
* fix(cost): price batch image completion tokens at image batch rate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): cover default vertex batch output transformation for image cost
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(cost): sync model prices schema
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(cost): simplify batch completion cost with rate helpers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(cost): keep global batch pricing fallback for image-only deployment rates
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(test): rename connect tunnel helper to https redirect
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(cost): merge deployment batch rates over global pricing
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(model_prices): put flash-image batch image rate on the GA row
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* chore(cost): drop redundant comment in batch fallback
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(cost): use _batch_or_half for the batch image rate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(cost): price nano banana 2.1 fixtures off the real cost map rows
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(cost): use the vertex 4k image token count in the batch cost test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): show Projects nav to team and org admins
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): satisfy frontend lint budget
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(ui): format leftnav regression tests
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(projects): let org admins see projects of every team in their orgs
/project/list and /project/info only knew proxy admins and team members, so an
org admin saw projects only for teams they had personally joined. Both now use
the same team access check as /team/info.
* fix(ui): one shared Projects access rule for nav, data and actions
The Projects query skipped global org_admin users, the page-visibility picker
could never offer Projects, and New/Edit showed to users the backend would
reject. Nav and queries now share one rule, Projects is selectable in the
allowlist, New/Edit follow team_admin_editable_team_fields, and the project
modal only lists teams the user administers.
* fix(projects): record IN-list bounds for project visibility filters
Both filters are bounded by one caller's team memberships and admin orgs. Also
cover a project whose team was deleted.
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: ryan-crabbe-berri <ryan@berri.ai>
* fix(logging): claim the async success dedup flag before the first await
Nested @client wrappers on the chat -> Responses bridge schedule two async
success tasks on one logging object. The dedup flag was checked at the top
of _async_success_handler_body but only set after several awaits; the
worker-thread base64 offload for messages >= 256 KiB yields there, so the
second task passed the same check and the request was logged twice
(two SpendLogs rows, key/team/daily spend counted twice).
Fixes#44500
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* fix(logging): serialise async success handlers per logging object instead of claiming early
Claiming the dedup flag before the first await (previous commit) let a first
handler that is cancelled before its callbacks run (e.g. by the logging
worker's per-coroutine timeout) suppress the second one, losing the request's
success accounting. Serialise non-streaming handlers on the same Logging
object with an asyncio.Lock kept in a module-level WeakKeyDictionary: the
second handler waits, then skips if the first logged, or logs itself if the
first raised or was cancelled first. Streaming chunk logging is not locked.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* test(integration): pin one spend row and one key charge for large bridged chat completions
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): wait for the large request's spend row before asserting
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(logging): use fixed timestamps in async success dedup tests
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(logging): trim the async success dedup lock comment
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(logging): reset background interaction dedupe under the success dedup lock
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(logging): use a fixed start time in the interactions logging helper
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix: claim the async success log at schedule time instead of locking the handler
Nested @client wrappers (chat over the Responses bridge, Anthropic Messages over the chat adapter) exit with one shared Logging object and each queued their own async success handler. The handler checked has_logged_async_success before its awaited work and set it after, so both could pass the check and the request was logged and billed twice.
_schedule_async_success_logging now claims the log on the Logging object synchronously through claim_async_success_log, and a later wrapper returns without enqueuing. There is no await between the check and the decision, so no lock is needed. The per-object asyncio.Lock, the WeakKeyDictionary that held it and the locked block in async_log_background_interaction_completion are gone.
Tests: the regression test moves to tests/unit/test_utils.py and drives two _dispatch_success_logging exits with one Logging object through the logging worker, asserting one log with the inner result. It fails without the claim. The cancelled-handler test is dropped since a second wrapper no longer retries. The background interaction completion test stays.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(logging): pin the background completion event order and name the innermost wrapper in the claim docstrings
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): send one bridged request and assert the inner Responses result is the logged one
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(lens): use async-timeout on python 3.10 for budget reservation timeouts
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* chore(deps): keep uv.lock diff to the async-timeout entry
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(lens): cover real request deadline expiry in reserved_budget
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(lens): finish Python 3.10 timeout coverage and dependency checks
* test(lens): control event-loop time for deadline regressions
* test(tracing): include priced call count in trace fixture
* fix(lens): limit timeout compatibility changes to PR scope
---------
Co-authored-by: Moe Khalil <moe@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(caching): never store or serve a chat completion with no choices
A provider response with empty choices was written to the response cache and served on every identical request until the TTL ended, with no provider call in between. The cache now skips storing such a response and treats an already stored one as a miss, so the next request goes back to the provider and its answer replaces the entry.
* fix(caching): skip responses with no output on the Responses API and Anthropic Messages too
* fix(caching): skip streams and stored entries that carry no output
A chat or text completion stream whose chunks carried no choice is closed
by the stream wrapper with one empty choice of its own, so the assembled
response passed the choices check and was cached. The assembled stream is
now judged on its content: a stream with no text, tool call, or other
output in any choice is never stored, on the async and sync writers alike.
The Responses API stream writer and the Anthropic Messages stream writer
apply the same no-output check before storing.
A stored entry with no output read through the worker memory tier is now
evicted from that tier on the miss, so the next read reaches Redis where
the refill lands; the text completion and messages writers only write to
Redis, and the memory copy otherwise kept missing until its own TTL.
* test(integration): response cache cells for answers without output
Deterministic cells for the response cache on every unified endpoint,
streamed and not, through the OpenAI and Anthropic SDKs and raw httpx,
plus the sync SDK paths, stale entries, malformed answers, per-request
TTLs, cache delete, and chaos (Redis stopped or paused mid burst, a
worker killed, in-memory cache mode). The scripted upstream counts only
POSTs as deployment calls, since the proxy's boot-time GET /v1/models
discovery of a config deployment is not one.
* test(caching): pin the stored entry timestamp in the worker-copy test
* test(integration): drop the restating comments from the chaos cells
* test(integration): close the breaker on the first call after the Redis restart
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(lens): price agent traces by joining gen_ai.response.id to spend logs
Trace spend now joins each model call to spend_logs on one key: the
span's response id (gen_ai.response.id or the id the normalizers read
from OpenInference/LangChain output) against spend_logs.response_id or
the upstream id embedded in a managed resp_ id. The litellm.call_id and
traceparent transport join paths and the per-row ownership gate are
removed; the spend SQL still restricts rows to what the reader can see.
A run with some unpriced calls now reports the sum of its priced calls
plus priced_calls, instead of an unknown total.
* test(lens): cover response id spend join and partial trace totals
* chore(lens): regenerate trace types for priced_calls
* feat(lens): show partial run cost as a lower bound with priced call count
* fix(lens): treat litellm.call_id as the same assigned call id for spend joins
The id LiteLLM assigned to a call is either the response id it returned
(gen_ai.response.id -> spend_logs.response_id) or its gateway call id
(litellm.call_id -> spend_logs.litellm_call_id). Both are exact ids the
gateway mints and logs, so the join stays one rule. Transport span
matching and the ownership gate stay removed.
* test(lens): cover litellm.call_id spend joins and restore captured totals
* feat(lens): link each priced model call to its spend log
Spans gain spend_log_request_id, the spend_logs.request_id the call was
priced from, and spend_match, which says whether a model call matched or
why not (no assigned id on the span, no spend log with that id, or an
ambiguous match). A model call span is priced from the same ids as the
run total, so its cost and the total agree.
* test(lens): cover spend log links on model call spans
* chore(lens): regenerate trace types for spend log links
* feat(lens): open the matched spend log from an LLM step
An LLM step's header now shows a Spend log chip with the matched
request id and cost; clicking it opens the request log drawer over the
run, fetched by the exact spend_logs.request_id instead of the span's
own response id. Unpriced steps say why (no assigned id on the span, or
no spend log with it). Tree rows show each model call's cost, and a
partial run cost shows its priced call count inline.
* test(lens): cover the spend log link and unmatched cost reasons
* feat(lens): show the spend log link as a bordered LiteLLM Spend Log button
* feat(lens): add a back link from the spend log drawer to the agent trace
* feat(lens): label the spend log back link Back to Lens trace with the Lens icon
* fix(lens): ignore assigned ids that name no spend log when pricing a call
An id that names no row no longer vetoes the call, so a span carrying
both a response id and a call id still prices from a spend row logged
before litellm_call_id existed. An id naming two or more rows makes the
call ambiguous, and the match reason comes from the same per-id result,
so a single matched row with no cost is reported as matched.
* fix(lens): price a trace only from spend logs in its own team
A reader with several teams could see the same assigned id in another
team's spend log; only rows from the trace's team now price it. The
user and key ownership gate stays removed.
* perf(lens): resolve each model call's spend once per trace
Model call matches are computed once when the trace is resolved and
looked up by span index, instead of scanning the model call list for
every span and walking the graph again for spans, agents and the run
total.
* fix(lens): hide a step's Cost fact only when its spend log link shows the cost
* chore(lens): drop narrative doc comments from the spend join
* fix(lens): price a model call only when its ids agree on one spend log per span
* fix(lens): keep pricing spend logs written before litellm_call_id by their request id
* fix(lens): price every attempt a model call's ids name when they agree
* feat(proxy): add opt-in vector_store_deny_by_default for standalone virtual keys
Adds general_settings.vector_store_deny_by_default (typed bool, default false). When enabled, a virtual key
with no team must list the requested vector store in its object_permission.vector_stores; no permission
record, null, or an empty list is denied with key_vector_store_access_denied. Omitted or false keeps the
existing behavior, including nonempty allowlist enforcement. The master key is unchanged in both modes.
Team keys and keyless callers are deferred to later increments of LIT-6035
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(proxy): require key and team vector store grants for team keys under vector_store_deny_by_default
With the flag enabled, a virtual key on a team needs both its own grant and its team's grant for every requested vector store. A missing permission record, an empty list or an unresolved team grants nothing. Dashboard session keys and the master key keep their existing behavior, and flag-off behavior is unchanged
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(proxy): require user or team vector store grants for keyless requests under vector_store_deny_by_default
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): cover vector_store_deny_by_default through a real proxy with key, team and JWT identities
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* chore(ui): regenerate schema.d.ts for vector_store_deny_by_default
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(auth): cover path and file_search vector store ids under vector_store_deny_by_default
Strict mode now reads vector_store_ids and tools[].vector_store_ids from the request body without needing a vector store registry, so /v1/vector_stores/{id}/search and Responses file_search are checked. User grants load through the object permission cache, and the proxy admin user rebuild keeps object_permission_id
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): broadcast user entitlement cache eviction to every worker
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(auth): reuse VectorStoreRegistry id extraction for vector_store_deny_by_default
Strict mode now uses get_vector_store_ids_to_run on an empty registry when none is loaded, instead of a parallel set of request-shape helpers, and vector_store_access_check documents the policy
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): add strict vector store audit cells for routes, SDKs, workers and concurrency
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): consolidate vector_store_deny_by_default coverage to core cases
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): reject invalid vector_store_deny_by_default at config load and return 400 for malformed vector_store_ids
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(auth): read vector store key and team grants through the object permission cache
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(auth): return typed permission rows in request flow vector store tests
Co-authored-by: mrinal <mrinal@berri.ai>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(auth): validate vector store ids and tools as immutable sequences
Co-authored-by: mrinal <mrinal@berri.ai>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: mrinal <mrinal@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(rust_bridge): remove rule-gated native secret-manager selection
Mirror the cache treatment: SecretManagerRule/SecretManagerContext and the resolve_native_* binding plumbing are gone. The Rust bridge now selects the native backend from the explicitly configured client (capture_secret_manager / _SecretManagerRuntime.from_client), and get_secret_from_manager is the plain Python handler path.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* chore(deps): bump sharp to 0.35.5 for GHSA-wq5f-xc86-pv6w
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Yujong Lee <yujong@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): budget s3 dedupe retries on the deployment so the seeded router num_retries cannot zero them
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(responses): prevent nested router retries
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): fix provider retry and pytest collection setup
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(responses): keep sdk retries on the sync router responses path
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(utils): count upstream calls instead of doubling the retry helpers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: mateo <mateo@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(types): replace Any with proven types in 299 files
Clears 871 basedpyright Any errors (reportAny 6,703 to 6,240, reportExplicitAny 1,651 to 1,243) without adding a cast, an ignore or a suppression, and without touching any budget file
Most edits are annotation-only: a parameter, return or local goes from Any to object, Mapping[str, object] or the concrete type the value always held. Fourteen files validate untyped JSON once where it enters, through a module-level pydantic TypeAdapter or model_validate, and then use real types
No HTTP status, error type or response shape changes. Mistral speech and fal.ai Bria image generation now report a pydantic ValidationError instead of an AttributeError when the provider answers 2xx with a body that is not a JSON object
* refactor(types): make the config locals fix effective and trim no-op edits
The Mapping[str, object] annotation on `locals().copy()` removed no error,
because the checker narrows the variable back to the dict[str, Any] the call
returns. 21 provider config constructors now build the same copy with
dict(locals()), which the checker infers as object values under that
annotation, so each file loses one reportAny.
The same annotation is reverted in 31 other config files where it stayed a
no-op, together with the tests that were added only to cover those lines, and
the one MCP server manager line that no CI coverage shard executes is
reverted too. The pull request drops from 345 to 297 changed files.
* refactor(types): accept only int in the proxy state setter
get_proxy_state_variable is annotated to return int, but
set_proxy_state_variable still took Any, so the checker could not hold
callers to the type the getter promises. The setter now takes int, which is
what its only caller already passes.
* refactor(types): index the proxy state key so the getter returns int
* refactor(types): keep public annotations and provider error text unchanged
Restore every public return, public method parameter, public attribute and exported
alias to its annotation on main so code that type-checks against the package keeps
type-checking, and take the Mistral speech and fal.ai Bria changes back out so no
provider error message differs from main
* refactor(types): leave the Vertex RAG chunking read as it is on main
Take the chunking format validation back out of the Vertex RAG ingestion path. It needs the vertexai SDK, a storage bucket and a RAG corpus to execute, so nothing here could run it end to end, and it cleared only two errors
* test(integration): pin the validated provider boundaries on a live proxy
* test(integration): give the held burst a client that outlasts the gate
The fault cell holds a burst at the upstream for up to 60 seconds while it kills a worker, but sent the burst through the shared 15 second client, so a slow box could time the survivors out before the gate opened. The burst now goes through its own client whose timeout is twice the gate, and the gate length is one named constant.
* refactor(types): keep the license reply handling and experimental MCP signatures as they were
The license check validated the whole reply as a mapping, which changed the error text logged for a reply that is not an object. It now validates only the verify value, so every reply is handled and logged exactly as before while the value is still typed.
Three files under the experimental MCP server changed annotations on public functions and methods (three returns and three parameters). They go back to their previous content so no public signature in the diff is narrowed.
* test(integration): answer the proxy's model-list call in the OpenAI stand-ins
Every 300 seconds each proxy worker asks an OpenAI deployment for GET /v1/models. Four new cells own an OpenAI stand-in that accepted only the call under test, so a refresh landing inside a cell failed it. The stand-ins now answer that call through the suite's own helper and the cells count only the provider calls they drive.
* refactor(rust): rename litellm-core to litellm-inference
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(rust): expose inference base API for format crates
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(rust): move shared inference test helpers behind test-support
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Yujong Lee <yujong@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(ui): wait for the classifier model popup before picking its option
Base UI exposes its select option asynchronously, and the helper waits for the option and its positioner to become clickable before selection
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: isolate two router tests from work leaked by earlier tests
Collect earlier tests' garbage before warning capture so their unawaited coroutines cannot be attributed to the target's warning assertion
Filter success-event callbacks by the request's litellm_call_id so queued logging work cannot replace the current request's captured messages
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor: remove fresh tech debt from the 2026-10-05 window
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(lens): drop the review models left unused by the dead review helpers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): price a model group from the deployments that serve it
A model group's info read composed the group's own model_group_alias
entry into its deployments, a hop the router never takes: an alias is
resolved exactly once at request time, so a group reached as an alias
target is served by its own deployments. For the chain X -> T -> U the
price read for X included U's deployments too, and the free-model
budget waiver refused a free request to X on an over-budget key, while
GET /model_group/info reported U's providers and price for X.
The group info read now prices a group from the deployments routing
serves it with: the ones named after it, the routing group of that
name, or the wildcard route matching it when neither exists. The
budget waiver, GET /model_group/info, the rate limiters, and the
response headers all read the same set as routing. get_model_list
keeps its behavior for every other caller.
* test(router): give the paid fixtures explicit per-token prices
* test(router): call the routed-group read by name so the router coverage gate sees it
* test(integration): audit the alias chain budget waiver on every route, shape, and outage
Thirty-four cells under the management group prove an over-budget key is served through an alias chain entry at the price of the deployment that serves it, on chat, responses, and messages, sync and streamed, through the OpenAI and Anthropic SDKs and raw httpx on both replicas, with the chain middle, the reverse chain, a cost-map priced middle, a ghost middle, wildcard and routing-group targets, malformed and hostile inputs, a cached reply, a repointed alias, a provider failure, and two chaos bursts (a killed worker, a scripted outage)
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* perf(types): defer pydantic schema builds via shared LiteLLMBaseModel
Add LiteLLMBaseModel with defer_build driven by DEFER_PYDANTIC_BUILD (default true) and move litellm and enterprise pydantic models onto it
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(types): build deferred models created by a parent validator; keep lens worker models litellm-free
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* chore(types): link pydantic issue on deferred-build rebuild hook
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: nate <nate@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
