* fix(realtime): keep turn_detection.type when a transcript guardrail disables auto-response
* test(realtime): type the GA session.update helper and mark its locals Final
* test(realtime): expect typed turn_detection on the voice guardrail session updates
* test(realtime): accept either gate path for the client's first voice session.update
* fix(gemini): drop VAD tuning when the guardrail turns auto-response off
* fix(gemini): check create_response without a redundant isinstance
* fix(vertex_ai): keep the client's VAD tuning on Vertex when auto-response is off
* fix(realtime): keep the transcript-guardrail gate typeless on xAI so it does not auto-answer
* fix(realtime): replace any client turn_detection with the untyped gate on xAI
xAI answers on its own under a typed server_vad even with create_response off,
so a client's typed turn_detection must not reach xAI while the transcript
guardrail gates the turn.
* fix(realtime): keep a client's early turn_detection and hide the guardrail cancel's error
A session.update sent before session.created reached OpenAI ahead of the proxy's own gate, so the gate replaced the client's turn_detection. The proxy now re-sends the client's gated turn_detection after the gate when the two differ.
The guardrail's response.cancel now carries a litellm_guardrail_ event_id, and the error OpenAI returns for it (response_cancel_not_active when nothing is answering) no longer reaches the client.
* fix(realtime): send the guardrail gate before a client's early session.update and let Muse and Chirp fall back to server_vad under drop_params
* fix(realtime): hide only the cancel errors that echo the guardrail's own response.cancel ids
Track the event ids of the guardrail cancels this session sent and drop a backend error only when its
error.event_id is one of them, so a client's own event id with the same prefix still gets its error.
Cover the scripted response.cancel reply in the integration upstream, update the stale list-mode test
to the typed gate, and mark the xAI cancel error cell as an expected failure, since xAI replies with
its own event id.
* test(integration support): deliver the disconnect a real server sends after the client goes away
The scripted upstream now keeps reading frames after a reply hits a closed socket, so the unit test's
fake ASGI receive has to end with the websocket.disconnect message a real server delivers.
* fix(realtime): delete the blocked transcript item before the guardrail's refusal turn
OpenAI GA answered the blocked question in some refusal turns because the
blocked user item stayed in the conversation. The raw OpenAI-protocol path
now sends conversation.item.delete for it right after the cancel. Provider
config backends have no such event, so that path no longer passes an item id.
Also bound the guardrail integration file by the shared graceful-stop
timeout and make the xAI cancel-error xfail require an AssertionError.
* fix(realtime): keep the blocked item on xAI, which cannot delete an audio input item
* test(realtime): assert the xAI voice block sends no conversation.item.delete upstream
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* feat(gemini): continue MAX_TOKENS decodes with continuationToken across chat, streaming, and interactions
* fix(gemini): continue long decode on CONTINUATION and cancel interactions at /cancel
* feat(gemini): add public gemini-4-argon cost-map rows for AI Studio and Vertex AI
Add gemini/gemini-4-argon and vertex_ai/gemini-4-argon alongside the existing
early-access row, carrying the same limits and capabilities. The AI Studio row
mirrors the early-access entry; the Vertex row differs only in provider and
pricing source. Prices follow once Google publishes them.
Extend the mapped metadata test to cover all three ids, that each routes to its
provider, that its backup copy matches, and that the public rows carry the same
limits as the early-access row.
* feat(gemini): raise gemini-4-argon to the 1M long decode cap and report initial_prompt_tokens across hops
* chore(prices): build on the merged gemini-4-argon rows and only raise the long decode cap
* fix(gemini): follow continuation hops in a loop instead of recursion
The unary hop collectors and the stream iterator called themselves once per continuation request, which the recursion check in CI rejects and which would hit the interpreter's call depth on a very long output. They now loop, and a test follows more hops than the interpreter allows nested calls
* fix(gemini): bill Google Search grounding from every continuation request
Google sends search grounding once, on the request that ran the search, and does not repeat it on a continuation request. The merged response kept only the last request's grounding, so usage.web_search_requests and the per-search fee dropped every earlier search. Unary merges and the stream iterator now carry the grounding of every request in the chain
* fix(gemini): continue long gemini-4-argon answers on streams and Vertex AI
A continued stream stopped after the first chunk of its second request. The stream wrapper asks the iterator for a fresh start before every chunk, and that reset it to the first request's finished stream. The iterator now reads from the newest request's stream
On Vertex AI the continuation request went to the v1 API version, which rejects continuationToken with a 400. A request that carries the token now goes to v1beta1, which accepts it, including a token that v1 issued
On a stream, initial_prompt_tokens came from a usage chunk in the middle of the stream, which reports a smaller prompt count than the last chunk. It is now the count the first request's stream ends with, the same number a unary call reports
continuation_token is no longer forwarded on the Interactions API, which answers 400 for the field. It is dropped there as it was before this branch
* chore(gemini): name the type errors the dual-mode stream iterator trips
The stream iterator holds a sync or an async line iterator in one untyped attribute. Pointing it at a continuation request's stream gave the type checker a concrete union for it, so six existing lines that only ever run in one of the two modes started to report errors
This adds a rule-named suppression with a reason on each of those lines and changes no code. Splitting the class into a sync and an async iterator would remove them and is left for a follow-up
* fix(gemini): continue only when Google ends a request with CONTINUATION, and stop streaming mid-answer tokens
A long decode now sends the next request only when Google's finish reason is CONTINUATION, so an answer cut at the caller's own max_tokens returns in one request with its continuation_token instead of drawing a rejected extra request
A request for more than one choice is never continued, since Google rejects a continuation with more than one candidate
Chat-format streams no longer carry the tokens Google sends mid-answer. Only a stream that ends cut off carries a continuation_token, on its last content chunk
The response cache key now includes continuation_token, so a resend with a token is not served the earlier segment
* fix(gemini): keep one response id across a continued answer
A continued chat stream changed its chunk id each time the next request to Google started, and a unary answer came back under the last request's id. Every chunk and the unary answer now carry the first request's id, including the chunk that ends a cut-off stream
A request whose n or max_tokens arrives as a string is sent once and never continued, where it used to fail with a 500 on the comparison. An empty continuation_token no longer changes the cache key
* test(gemini): add integration cells for long decode continuation
Covers continued answers on chat, /v1/messages and /v1/responses, streamed and not, on the Gemini API and Vertex AI against a scripted Google, plus concurrency, outage and worker restart runs
Also moves the Google stand-in cancel route to /cancel, the URL this PR now calls for Interactions cancel, and checks the cancel test posts there exactly once
* fix(vertex_ai): fetch the access token again before each continuation request
A long Gemini 4 answer on Vertex AI can run for over half an hour of follow-up requests, and every one of them reused the bearer token fetched before the first request. Once that token expired Google answered 401 and the caller got a cut-off answer with finish_reason length
Each Google request in a chain now asks for the token again, which is a cache hit until the token nears expiry, on the streamed and unary paths, sync and async. A refresh that fails counts as a failed follow-up, so the caller still gets the answer so far with its continuation token. A caller's own Authorization in extra_headers still wins on every request, and the Gemini API key path is unchanged
* fix(vertex_ai): start each continuation stream with an empty parser buffer
When a Google stream ended on an unfinished frame, its leftover text stayed in the stream parser and every frame of the next continuation stream was appended to it. None of them could be decoded, so the whole continuation was billed and dropped without an error
Each continuation stream now starts with a fresh buffer, and a test covers a stream that ends on an unfinished frame, sync and async
* fix(vertex_ai): treat a continuation stream with no readable frame as a failed follow-up
A continuation stream that answered 200 and then sent Google's error frame first, sent a body that is not JSON, or dropped before its first frame ended the caller's stream with `stop` or an error and no token, so a cut answer looked complete or was lost. It now ends the way a refused follow-up does, with the answer so far, `finish_reason` length and that request's token
Every continuation stream request now also gets the caller's per-request timeout, which #46349 added to the first one, so a follow-up that never answers ends at that timeout the same way
* fix(vertex_ai): keep every follow-up's citations and URL context in a continued answer
A non-streamed answer that took several Google requests kept citationMetadata and urlContextMetadata from the last request only, and left every later request's offsets counting from that request's own text. So a URL read on the first request was missing from the answer, citations from earlier requests were dropped, and a later request's citations and search supports pointed at the wrong part of the joined answer
The joined answer now keeps every request's citations (Vertex AI's `citations` and AI Studio's `citationSources`) and URL context, and shifts each later request's citation and search support offsets by the UTF-8 bytes of the answer text sent before it, thought parts excluded, since Google counts these offsets in bytes
* fix(vertex_ai): answer when a candidate sends null citationMetadata or urlContextMetadata
A Gemini candidate whose citationMetadata or urlContextMetadata was JSON null made the call fail on every Gemini model, non-streamed with a 422 or 500 and streamed with an error event, because the per-request metadata helper handed the null back as is and the caller tried to extend a list with it. A continued answer also failed when a follow-up's citation or search support came with a null startIndex
A null field is now kept as one entry on a single request, as it was before this branch, a follow-up's null URL context adds nothing to a continued answer, and a null startIndex counts from the start of that request's text. Tests cover missing, null and empty fields on one request and on a continued answer, streamed and not
* fix(vertex_ai): pass through spans with no end index and point a continued stream's citations into the whole answer
A later request's citation or search support with no endIndex made the whole non-streamed answer fail with a 500, so a span Google sends without an end index now comes back as Google sent it, the way one request for the whole answer returns it. A span with an end index but no start index still counts from 0
A continued stream handed callers each later request's citation and search support offsets counted from that request's own text. The stream now adds the answer bytes of every earlier request, so the offsets point into the whole answer like the non-streamed path and one request for the whole answer
Adds the audit's wire tests for both, which fail 24 cells before this change
* test(vertex_ai): cover a later request whose cited span starts at its first byte with no start index
Google leaves out a zero startIndex, so the second request of a continued answer can cite its own opening with only an endIndex. These wire cells check that span lands where one request for the whole answer puts it, on Gemini and Vertex AI, chat and Responses, unary and streamed
* fix(vertex_ai): log a continued answer's raw response as one object
A non-streamed answer joined from several Google requests logged its raw response as a JSON list of every request's body. The OpenTelemetry logger reads that raw response as one Gemini body to set its `llm.gemini.*` attributes, so on a continued answer it hit a validation error and set none of them
It now logs the merged body, which carries only the last request's token, and a single request still logs Google's body unchanged. The new unit test fails with the change reverted
Also adds integration cells for a cut chat stream's token chunk, an answer joined from 41 requests, a caller that gave up on a non-streamed call, counts and caps that are not numbers or are sent twice, `n` sent as true, a partial answer cached after a refused continuation, the token in another provider's cache key, and the sync SDK path with its Vertex token refetch
* test(integration): check a continued answer logs one merged raw response
A non-streamed Gemini answer that took several Google requests now hands logging one merged response, so the OpenTelemetry span carries the llm.gemini attributes. These 12 cells fail on the merge base and on the commit before that fix, and pass on the branch
* test(vertex_ai): cover a follow-up tool call and its thought signature in long decode
* test(gemini): pin that a one-request answer logs Google's body unchanged
* test(integration): isolate every multi-worker integration suite by default
Per-worker database, Redis, upstream and proxy were opt-in for management,
accounting, providers and extensions. mcp, security and cost ran 4 to 8
xdist workers against one shared proxy, database and Redis, and those suites
hold most of the integration flakes CircleCI records. Every run with more than
one worker now builds a private stack per worker, run_integration.sh no longer
starts or exports the shared proxy, upstream, database or Redis in that mode,
and each worker's scripted upstream runs in its own process instead of a
thread in the pytest controller.
Worker gateways get what the shared proxy used to provide: a peer for
management and mcp, the scripted cost map for cost, stdio MCP servers, and
coverage for mcp. mcp coverage now measures lines through sys.monitoring,
because branch measurement on CPython 3.12 uses the trace function and that
made a covered proxy take about 4x as long to become ready, which is why the
shared mcp proxy kept missing its 90 s readiness wait. Routing parity
snapshots now also read the worker databases, since its cost and mcp runs use
several workers. The security per-worker fixture is redundant and removed
* test(integration): give the nested probe root the files a worker gateway needs
Two worker probes now run isolated, so the nested controller boots a real per-worker upstream and proxy from the probe root. That upstream imports the fake OpenAI server and the cost map modules, and the proxy reads proxy_config.yaml, none of which the probe copied. The nested run also drops INTEGRATION_COVERAGE_FILE so it never writes into an outer coverage file
* test(integration): budget nested multi-worker runs for their worker stack boots
A nested run.py database run with more than one worker now boots a private database, Redis, upstream and proxy per worker, one after another, before any case runs. The fixed 60 s subprocess timeout left no room for that, so the 2 worker probes sat at 57-60 s on CircleCI. The subprocess budget is now the old 60 s plus 30 s per worker stack, and the outer pytest timeout keeps its previous 30 s of slack on top. 30 s is the slowest worker proxy readiness recorded in startup.json on the integration-cost job of workflow e09b108b (21.6-30.3 s on 4 CPUs, 2026-10-12)
---------
Co-authored-by: Yuneng Jiang <yuneng@berri.ai>
* test(integration): wait for customer spend rows and use owned OpenRouter cost
* ci: isolate integration workers and clean up resources after crashes
* docs: describe isolated extensions workers and shard census
* fix(ci): restore integration service and fixture guarantees
* fix(integration): preserve replay IDs and native route fixtures
* fix(tests): scope integration harness imports to each unit case
* fix(ci): preserve fixtures and wait for owned workers
* fix(ci): isolate nested runs and telemetry cases
* fix(ci): make worker readiness and fault tests deterministic
* chore(ci): drop MCP fixture repair already covered by main
* chore(ci): keep the local catalog fix from main
---------
Co-authored-by: yuneng <yuneng@berri.ai>
* test: move guardrails, image, audio, batches, search legacy tests to unit
* test: restore the custom guardrail and remaining chunk 1 unit rows
* test: move the ocr matrix to e2e
* test: drop comments from moved tests and fix vertex ocr credentials
* test: move litellm utils and unified google legacy tests to their tiers
* test: move logging callback and pass-through unit legacy tests to their tiers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: move store_model_in_db legacy tests and start the admin ui migration
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: finish the proxy job rows and drop their CircleCI jobs
Moves the remaining multi-instance, openai endpoint, otel, pass-through and
top-level proxy tests into tests/integration with scripted upstreams, rewrites
the weak replacements to assert the outbound wire and exact responses, and
removes the build_and_test job group this PR owns from CircleCI
* test: tighten the key generation and error log assertions, rename migrated unit files to their mapped names
* test: shard the pagerduty unit tests and record the rust harness file on the coverage allowlist
* test: drive the SSO sign-in tests through a scripted IdP and isolate the cost center stand-in
* test: type the e2e beta header, prompt cache and secret manager values
* test: assert the missing model error message on the OpenAI routes
* test: drop the deleted otel base class import and restore mcp redaction and exact langfuse usage checks
* test: make the langfuse fixture directory a package for the layout check
* test: restore collected no-match cases, real langfuse and gcs payload checks, and gate the aws secret lane
* test: keep the ocr e2e pdf fixture next to the test
* test: put meta below covers on the streaming tool search e2e
* test: run the OCR matrix against a scripted upstream and fix two runner-only e2e failures
* test: count only OCR calls in the OCR matrix recorder
* test: pin OCR parse bodies and billed cost, group new e2e cases into classes
* ci: restore the python image anchor dropped in the main sync
---------
Co-authored-by: yuneng <yuneng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): move the openai cost tracking cases from cost_tracking_cases.json to CostTrackingTestCase literals
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): move the remaining cost tracking cases, the cost map and the batch/realtime cases from cost_tracking_cases.json to Python literals (#45885)
Co-authored-by: kerry <kerry@berri.ai>
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>
* test: migrate local_testing SDK and provider tests out of build_and_test
* test: cover sagemaker region, hf template and nova paths end to end
Region rows now go through litellm.completion on the sagemaker/ handler that
the legacy tests exercised, the prompt template test mocks the Hugging Face
tokenizer config instead of switching to custom_prompt_dict, streaming tests
assert role, finish reason and body, and Nova single turn and streaming run
through litellm.completion with a recorded request body
* test: tighten migrated integration tests for local OpenAI routes, timeouts and air-gapped boot
Split the catch-all providers migration file by behavior. The fake endpoint
route tests now hit the owned upstream directly, as the legacy harness tests
did. LM Studio asserts the outbound key, path and stripped model. The embedding
timeout test asserts the proxy maps a deployment timeout to a 408 and forwards
the read timeout to the provider client, instead of timing out the test client
The Azure streaming tool-call test also asserts the second identical stream is
served from the Redis cache with the same reassembled tool call
The Docker network test always skipped in CI because the security job has no
litellm/litellm image and no egress to pull one, and it tested a published image
rather than the checkout. It now boots the checkout proxy with every outbound
HTTP connection refused and asserts readiness, the bundled cost map fallback and
that the cost map host is the only egress attempted
* test: tighten the migrated e2e rows for local_testing SDK and provider tests
Drop the duplicated Anthropic covers marker and the Anthropic cell wrongly claimed by
the Bedrock and Vertex cache tests. Keep the file-wide Anthropic model as it is on main.
Send the 2023-06-01 version header to Anthropic through the deployment's extra_headers
so the header actually reaches the provider, and mark the creation-token case 1h then
5m like the legacy test did
Reuse the Bedrock file's own registration for the pydantic response test, swap the
hand-built PNGs for the shared cat.jpg fixture, assert the SageMaker Nova image answer
names the cat and that top_logprobs=3 is honored, and register the SageMaker Nova opt-in
markers the same way as the other opt-in markers
* test: restore lost assertions in migrated test_main tests
Restore the role alternation check in test_completion_anthropic_hanging, the
embedding and streaming response-header checks, the logit_bias drop check and
the gemini system and safety settings checks. Add an azure 429 retry-after
test for test_exception_with_headers, trigger the router fallback with
mock_timeout like the legacy test, and stop the mock timeout tests sleeping 3s
* test: restore lost legacy assertions in migrated vertex ai unit tests
* test: tighten migrated cost, streaming and embedding tests and scope the hosted mock guard
* test: strengthen migrated anthropic, azure, bedrock and cohere provider tests
Assert full request bodies and usage instead of presence checks, restore the
dummy-tool and custom_llm_provider assertions the legacy tests made, add the
async and per-model variants the legacy parametrization covered, split
streamed tool arguments across deltas, and move the Azure chat streaming
tests to the azure.py mirror file
* test: tighten migrated llms provider tests and cover legacy intents lost in the move
Strengthen the migrated fireworks, gemini, mistral, together, openai text completion, openrouter, perplexity and openai error tests so they assert what the legacy tests proved (cost, provider tag, usage-only-on-last-chunk, function_call rebuild, reasoning param, sync/async/stream error fields). Replace the huggingface and jina tests that doubled litellm code with respx end-to-end tests. Add the tests the pre-existing mapped tests never covered: cohere 400 mapping, token-array text completion, hosted_vllm/llamafile/lm_studio env api base embeddings, watsonx embedding auth, ollama ssl_verify client and ollama vision prompt
* test: cover residual legacy paths for anthropic tool streams, watsonx async stream, ollama http images and caller async embedding clients
* test: restore single request assertions in the deep_infra and azure_ai embedding tests
* test: drop a no-op streaming cost flag from the openai streaming cost test
* test: cover residual legacy-test lines in caching, main, utils, replicate, vertex and bedrock invoke
* test: cover streamed audio assembly, choice membership, untranslated responses events and the stream log guard
* test: pin cost rates in unit tests, use sdk for cohere e2e, drop vertex llama e2e
* test: send the vertex maps e2e request through the openai sdk
* test: drop the local_testing jobs that select no tests after the migrations
* test: wait for the streamed cache entry and keep custom Mistral rates test-local
* test: keep untouched Mistral test lines as they were
---------
Co-authored-by: yuneng <yuneng@berri.ai>
* fix(vertex_ai): dial the multi-region Live API host for realtime sessions
A realtime deployment with vertex_location us or eu dialed
{location}-aiplatform.googleapis.com, which Vertex does not serve for a
multi-region, so the WebSocket handshake came back 404. The URL builder now
resolves its host through the shared Vertex host resolver, which already
maps us and eu to aiplatform.{geo}.rep.googleapis.com and leaves regions
and global unchanged.
* test(vertex_ai): annotate the realtime multi-region URL tests
* fix(vertex_ai): skip the TLS argument when the realtime api_base is plain ws://
The Vertex realtime session and health-check connects passed the shared
TLS context to every websocket dial, so an api_base the transformation
already maps from http:// to ws:// failed with 'ssl argument is
incompatible with a ws:// URI' before any frame was sent. The OpenAI
realtime handler already skips the argument for ws://; the shared helper
now does the same for the Vertex path
* test(integration): cover the Vertex realtime multi-region host through the proxy
Fourteen cells on the scripted Gemini Live upstream: api_base overrides for
us, eu, us-central1 and global complete a turn; malformed locations are
refused before any dial; a missing location defaults to us-central1; the
realtime health check handshakes the override host and reports a malformed
location without dialing; a refused handshake reaches the client and the
next session connects; twelve concurrent sessions across locations each
reach the upstream once; an upstream outage closes every open session and
the proxy recovers
* test(realtime): type the health-check helpers and flatten the burst order
* test(http_handler): type the realtime TLS test signatures
* fix(realtime): skip guardrail VAD session.update injection for transcription sessions
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(realtime): flag transcription sessions from the route intent and backend events only
A client session.update declaring session.type transcription on a voice
session no longer sets the transcription flag, so it cannot switch off the
guardrail's create_response gate or skip the transcript guardrail
* fix(realtime): flag transcription sessions from provider-transformed session events
* test(integration): cover transcription sessions skipping the VAD auto-response injection
Adds the realtime transcript guardrail audit cells: transcription sessions on all three
realtime routes, the OpenAI SDK, the beta protocol, the whisper default deployment, the Azure
GA path over a TLS scripted upstream, and Meta Muse push-to-talk sessions keep the client's
session.update verbatim and get their transcript, while voice sessions keep the injected
create_response gate and a client-declared transcription type no longer bypasses it. Sad,
edge, and chaos cells cover malformed session fields, duplicate and older backend session
events, unauthenticated upgrades, repeated sessions, an upstream outage under open sessions,
and a worker kill with a proxy restart.
The scripted upstream now answers session.update the way the vendor does (session.updated,
or the missing turn_detection.type and session-type errors), records every websocket frame,
serves the Azure and Muse realtime paths, and can run over TLS from an owned upstream.
---------
Co-authored-by: gabriele <gabriele@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(openai/realtime): drop model from upstream URL for intent=transcription
* fix(openai/realtime): keep model on the upstream URL for non-OpenAI hosts
A custom api_base on the openai provider can be a gateway that routes on
the model query param, so the transcription model drop now applies only
to OpenAI's own hosts. xAI's URL is unchanged again
* test(integration): audit realtime transcription upstream URL on a custom api_base
Twenty-two integration cells cover transcription and conversation sessions on every realtime route, the OpenAI SDK sync and async clients, malformed and duplicated query params, unauthenticated and refused upgrades, an unknown model, idempotent spend logging, a twenty-session burst, an upstream outage mid burst, a worker kill, and a proxy restart, each asserting the exact query pairs the upstream received. The scripted upstream now records every websocket upgrade as an observation
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(proxy): return 400 instead of 500 for missing required params and invalid pagination
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci: run search_endpoints tests in proxy-endpoints shard
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): return 4xx for missing required params across all LLM routes and propagate provider status on lookups
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(llm_http_handler): keep provider error text when re-raising mapped errors
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): allow promptless image edits and default search models
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): default missing image edit image to None
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(proxy): build image edit defaults without mutating request data
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): keep image edit defaults within type-discipline budget and give request mocks a scope
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): inject a fake router for the search default model test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(llms): cover provider error status on vector store and file lookup handlers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(llms): keep the lookup handler raise block to a single statement
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(llms): cover provider error status on eval, eval run, skill and vector store file content lookups
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): cover missing required body params and provider lookup status codes
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci: run tests/unit/proxy/search_endpoints in the proxy-endpoints shard
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): bind spend-row request id with partial to satisfy B023
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): only reject non-positive page_size on vector store list
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): remove unreachable fine-tuning body validation
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): cover streaming anthropic messages reaching the upstream
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): count only provider calls when asserting missing params never reach the upstream
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): preserve merge-base request compatibility
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): preserve interaction completion model defaults
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): retry model read-through before rejecting params a DB-only deployment may default
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: shivam <shivam@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: yucheng <yucheng@berri.ai>
* feat(interactions): durable cross-pod settlement for background interaction billing
Background interaction billing lived only in the creating replica's memory, so a
DELETE routed to another replica, or a restart of the creating one, never billed
the completed provider work and the budget reservation was refunded at the poll
timeout. The create now registers the billing context in a settlement store
before returning, the proxy installs a Prisma-backed store at boot
(LiteLLM_BackgroundInteractionSettlement, schema-only migration), any replica
claims the row once through a conditional update before billing or releasing,
startup resumes every unclaimed row with its remaining timeout, and a give-up
records an unsettled outcome instead of silently reconciling to zero. The SDK
keeps an in-memory store and behaves as before.
* fix(interactions): survive a settlement install failure at boot and stop carrying request headers
* fix(interactions): drop the stored request context once a settlement row is settled
* fix(interactions): bill the completed response a poll already saw when its claim only answers at the deadline
* fix(interactions): carry a missing model through the settlement context for agent-only background creates
An interaction created with an agent and no model reaches the poll with no model name, exactly as on main. The settlement context now stores that None instead of rejecting the create, which answered the client with a 500 after the provider had already accepted it.
* fix(interactions): leave an unfetchable background interaction to its creating poll when a delete lands elsewhere
The remote pre-delete path fetches with only the delete's credentials, so a fetch it cannot make says nothing about the interaction. It used to claim the settlement row and release the reservation anyway, which stopped the creating replica's poll and lost the bill when the delete then failed the same way. It now returns without claiming; the in-process path keeps releasing on an unfetchable state, since its context carries the create's own credentials.
* fix(interactions): fail a cross-replica delete when its pre-delete fetch fails so the creating poll keeps the bill
* fix(interactions): keep the stored settlement gate when registration raises after landing, and fail resumed-poll deletes closed
A registration that raised after its row committed moved the poll to a private in-memory gate, so the creating worker billed while the stored row stayed unclaimed for another replica's delete or the next boot to bill again. The row is now read back once and, when it landed, the poll claims through it like every other settler.
A worker that resumed the poll after a restart is not the creator, so its delete on a failed pre-delete fetch now fails with the fetch's error instead of releasing and deleting. After a fleet restart every worker holds resumed polls, which left the fail-closed path applying nowhere.
* fix(interactions): settle an unverified registration through the durable claim
A create whose settlement-store write raised no longer bills through a
private in-memory gate that a later boot's resume cannot see. The claim
asks the durable store first and falls back to the local gate only when
the store answers that no row exists, and a missing settlement table reads
as no rows so a replica without the migration still settles in process.
* test(proxy): keep the settlement test where the proxy-infra shard collects it
The merge of main moved test_background_interaction_settlement.py under
tests/unit/proxy/spend_tracking, but the proxy-db shards claim tests/unit/proxy
files one by one in .circleci/scripts/unit_selection.sh, so no CI shard ran it
and codecov/patch dropped. tests/test_litellm/proxy/spend_tracking is collected
whole by the proxy-infra shard, which is where the test ran before the merge.
* fix(interactions): raise on a non-2xx Gemini interaction fetch
AsyncHTTPHandler.get never raises for status and the Gemini GET transform
only raised when the body was not JSON, so a 500 or 404 carrying Gemini's
JSON error body parsed as an interaction with no status. A delete on a
replica other than the creator then claimed the settlement as released and
forwarded the delete instead of failing closed, and the bill was lost. The
transform now raises GeminiError with the vendor's status, as the delete
transform already does; the in-process poll already retries a fetch that
raises
* test(integration): audit durable background interaction settlement across replicas
Twenty-six deterministic cells drive a one-worker creator and a two-worker
settler against an owned scripted Gemini upstream: cross-replica deletes
bill once, failed and cancelled interactions release, a later replica
resumes unclaimed rows, custom deployment pricing bills at the deployment
rate, a fetch the settler cannot make fails the delete closed, odd ids are
refused, a missing settlement table keeps in-process billing, polling
disabled registers nothing, the budget reservation is released by the
settler, an upstream outage mid-burst fails closed and recovers, killed
workers hand their polls to the respawned ones, and concurrent deletes on a
slow upstream settle exactly once. The support upstream gains a scripted
interaction store with per-id GET status and delay, and the process helper
gains an owned upstream a test can stop and restart
* test(integration): refuse a repeated delete in the scripted upstream and pin the settlement budget below one estimate
* chore(ui): regenerate dashboard API types after merging main
* test(integration): accept the 422 budget refusal and a respawned worker's resume
The budget cell pinned a 400 that the proxy stopped answering when budget refusals moved to 422, so it now asserts the status and the budget_exceeded error type the sibling budget tests pin. The later-booting replica cell accepts a claimer that is any worker started after the creates, since uvicorn's supervisor can respawn the creator's worker under load and the respawned worker's boot resume claims the rows by design; the single spend row check is unchanged
* test(integration): delete the pinned key's interaction with a second key
A key whose budget is filled by its own reservation is refused on every route, the DELETE included, so the cell now asserts that 422 and sends the delete with a second key, which is what the reservation release on another replica needs in order to be observable at all
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* test(integration): reproduce encrypted_content_affinity 503 when origin has no boundary peer
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): keep encrypted content affinity turn one out of the response cache
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): degrade encrypted_content_affinity when the origin has no encryption-boundary peer
A routed-group candidate that is currently unavailable and shares its
(api_base, api_key) with no healthy deployment used to raise a proxy-level
503/429 from _unavailable_origin_error, even though healthy siblings in the
same model group could still serve the turn. Strip the encrypted reasoning
and dispatch to the healthy pool instead, matching the existing cross-group
behavior, and log a warning naming the origin model_id and routed group
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): bound the degraded-affinity log marker and assert the strip on the wire
Address review findings: restore num_retries alongside
optional_pre_call_checks in the integration test teardown, record scenario
request bodies on the scripted upstream so the tests can assert no encrypted
reasoning reaches the sibling, and truncate the client-supplied model_id in
the degraded-dispatch warning
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(tests): only record JSON bodies on the scripted upstream
Multipart uploads to scripted POST routes have no JSON body, so gate the
observation recording on the request content-type
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): make encrypted_content_affinity runtime-toggleable so /config/update can turn it off
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(vector_stores): keep config-defined vector stores listed and read-only
Vector stores declared in config.yaml were purged from the in-memory registry by /vector_store/list because the database was treated as the only source of truth. Config-defined stores now carry is_config=True, stay in the list beside database rows, are never overwritten or evicted by database state, and reject /vector_store/new, /vector_store/update and /vector_store/delete with 400. The Admin UI renders them read-only
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): show vector store source and read-only state for config-defined stores
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): audit config-owned vector stores across list, writes, search, authz, peers and redis outage
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): show a visible read-only hint in the config vector store actions menu
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: ryan <ryan@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Since #40994 the proxy relays the upstream error body on a 400, so the
public error type is now whatever the upstream sent instead of the
status-derived name. The fake upstream answered every scripted failure
with type api_error, which made the public-error contract in
test_retry_counts_and_public_errors_match_actual_provider_attempts fail
on main. The fake now sends the type a real OpenAI-compatible upstream
sends for the status, so the assertion holds whether the proxy relays
or maps the type
