* test: drop the cwd-relative sys.path.insert calls from the test suite
TQ003 stands at 1,077 across 1,058 files, and 1,015 of them are the same shape:
sys.path.insert(0, os.path.abspath("../..")) and its deeper siblings. The
argument resolves against the working directory rather than the file, so from
the repo root, where every job runs pytest, it inserts the directory two levels
above the checkout. It has never pointed at litellm. The package is installed
into the environment anyway, which is what actually makes the import work, and
what the rule's message has said all along.
Removing them leaves 1,634 imports of sys and os with no remaining reference,
and those go too, except where another test module imports the name back out of
the file. The rest of TQ003 is 62 call sites that resolve against __file__ or a
variable, which are a different question and are left alone.
Collection is identical either way: 45,871 tests and the same 51 pre-existing
collection errors before and after, and ruff reports no new undefined name.
* test: drop the duplicate imports the sys.path sweep exposed to F811
* test(pre-call-utils): restore the os import the new bedrock tests need
A test that asserts on the error inside its own except block passes when the
call stops raising, because nothing runs the handler. That is the exact case
the test exists to catch, so the regression lands green.
Rewrites all 111 such blocks into pytest.raises, which fails when the call
succeeds, and selects PT017 in ruff-tests.toml so no new one lands.
`pytest.raises(Exception)` with no `match=` passes on any error that broad. A
TypeError from a refactor, a botched fixture, an import that moved: all of them
read as the rejection the test claims to police, so the test goes green for the
wrong reason and stays green after the behaviour it guards is gone.
PT011 closes that gap for the 317 sites B017 could not reach, because B017 only
fires on a single-statement body with no `as e` binding. Each pattern here is the
message the code actually raised, recorded by running the sites under a plugin
that logged the concrete type and text per call site, so the assertions describe
observed behaviour rather than a guess. Where a site raises more than one message
across its parametrize cases, the pattern is an alternation of what was seen;
where the exception carries an empty `str()` and puts the text on `.message`, the
site keeps a narrow `noqa` with the reason.
PT014 removes four parametrize cases that were listed twice. The duplicate re-runs
an assertion that already passed, and it usually marks a case someone meant to
vary and forgot to edit.
* test: enforce PT012 so a pytest.raises block cannot hide dead assertions
`with pytest.raises(...)` stops at the first statement that raises. Anything
sequenced after it inside the block never runs, so an assertion written there is
never checked and the test still reports green.
Two sites were doing exactly that, and both assertions turned out to be wrong
once they started running. tests/llm_translation/test_prompt_factory.py asserted
the bedrock rejection names "requires at least one non-system message", which
holds. tests/proxy_unit_tests/test_proxy_server.py asserted the prisma startup
failure mentions "httpx.ConnectError", which never appears: the failure is an
httpx.ConnectError whose message is "All connection attempts failed", so that
test now asserts the type. Its DATABASE_URL override moves to monkeypatch, since
the old restore sat below the assertion and leaked the invalid URL into every
later DB test the moment the assertion started being able to fail.
The remaining 72 sites are rewritten without changing what they exercise: setup
that cannot raise moves above the block, a nested `patch` moves outside it, and
bodies with real control flow (a stream drain, an if/else on sync_mode, a
retry loop) move into a local closure the block calls.
Fixing PT012 unmasked two B017s, since ruff only reports a blind
pytest.raises(Exception) once the block holds a single statement.
tests/proxy_unit_tests/test_auth_checks.py narrows to the ProxyException
can_key_call_model actually raises. tests/local_testing/test_completion_cost.py
was asserting vertex_ai/medlm-medium has no cost entry, which stopped being true
at some point; that dead first half is gone and the rest of the test, which
checks medlm pricing resolves above zero, now runs instead of being skipped.
* chore(ci): ratchet TQ004 to 768 after the prisma test moved to monkeypatch
* fix(main): stop per-request custom pricing from clobbering shared model_cost pricing
A request routed through a wildcard deployment with explicit zero pricing
(e.g. openai/* with input_cost_per_token: 0) registered that pricing on the
shared {provider}/{model} key in litellm.model_cost, so sibling deployments
relying on built-in pricing logged $0 until process restart (LIT-3991).
Request-time registration in completion()/embedding() now mirrors the
router-startup isolation: router-originated requests register full pricing
under the deployment's unique model id only, while the shared backend key
receives the entry with custom pricing fields stripped. Direct SDK calls
without a router deployment id keep the legacy shared-key registration.
The stripping logic is shared via
CustomPricingLiteLLMParams.strip_custom_pricing_fields and reused by
Router._create_deployment and Router.add_deployment.
* test: update legacy tests that asserted per-request pricing leaking into shared model_cost
test_router_fallbacks_with_custom_model_costs asserted the shared
claude-sonnet-4-5-20250929 entry ends up with the deployment's 30/60
pricing, which is exactly the cross-deployment leak this PR removes; it
now asserts the shared key keeps the built-in pricing, matching the
test's stated goal.
test_cost_calc.py::test_run computed streaming cost via
completion_cost(response), which only matched the non-stream cost while
the shared gpt-3.5-turbo entry was poisoned with the per-request
2/token pricing; it now passes the request's custom pricing explicitly
via custom_cost_per_token.
* test: point router/completion/triton tests at the local fake OpenAI endpoint
The shared Railway-hosted mock (exampleopenaiendpoint-production.up.railway.app)
takes down unrelated CI jobs whenever it is unreachable. #30695 moved the mounted
proxy configs onto a job-local fake server but left these in-Python api_base
literals pointing at the dead host, so litellm_router_testing, local_testing_part1,
local_testing_part2 and llm_translation_testing still fail with a 404
"Application not found" when Railway is down
Resolve the api_base from FAKE_OPENAI_API_BASE (default http://127.0.0.1:8190)
through a shared helper, auto-start the canned server from the local_testing and
llm_translation conftests when nothing is already serving, and extend the server
with a Triton embeddings route and a slow-endpoint delay so the triton and
latency-timeout tests run fully offline. The deliberately broken fallback URL is
left as-is so fallback handling still has a failing upstream
* fix: ignore non-loopback FAKE_OPENAI_API_BASE so the local mock is used in CI
* fix: drop 0.0.0.0 from loopback hosts, an unreliable client connect target
* fix(tests): keep fake OpenAI mock alive across xdist workers
ensure_fake_openai_endpoint registered atexit on the worker that spawned
the subprocess, so under -n 4 the first worker to drain its queue would
terminate the shared mock while siblings were still hitting it. Detach
the child via start_new_session and drop the per-worker teardown; reuse
on /health handles re-runs and CI containers clean up themselves
The test was failing because it depended on real API calls to deprecated
models. Now uses mock_response to validate streaming through the router
without external dependencies.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- test_async_fallbacks_streaming: replace deprecated gpt-3.5-turbo fallback
with gpt-4o-mini, fix use of module-level kwargs variable
- test_ausage_based_routing_fallbacks: remove Redis dependency to prevent
shared state across parallel CI containers (test already uses mock_response)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- test_async_fallbacks, test_async_fallbacks_streaming, test_sync_fallbacks:
update previous_models assertion from 4 to 3 (fallback not counted)
- test_ausage_based_routing_fallbacks: update deprecated model
claude-3-5-haiku-20241022 to claude-haiku-4-5-20251001
- test_router_fallbacks_with_cooldowns_and_model_id: increase RPM from
1 to 2 so second request isn't blocked by RPM consumed during failed
first request
- test_sync_in_memory_spend_with_redis: add delay after constructing
RouterBudgetLimiting to let background init tasks complete before
overwriting Redis values
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- test_router_cooldown_handlers: add mock_response to avoid real API call requiring OPENAI_API_KEY
- test_router_timeout: update deprecated claude-3-5-haiku-20241022 to claude-haiku-4-5
- test_router_fallbacks: relax assertion from == 4 to >= 3 to handle cooldown timing variance
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat(main.py): initial commit for `/image/variations` endpoint support
* refactor(base_llm/): introduce new base llm base config for image variation endpoints
* refactor(openai/image_variations/transformation.py): implement openai image variation transformation handler
* fix: test
* feat(openai/): working openai `/image/variation` endpoint calls via sdk
* feat(topaz/): topaz sync image variation call support
Addresses https://github.com/BerriAI/litellm/issues/7593
'
* fix(topaz/transformation.py): fix linting errors
* fix(openai/image_variations/handler.py): fix passing json data
* fix(main.py): image_variation/
support async image variation route - `aimage_variation`
* fix(test_get_model_info.py): fix test
* fix: cleanup unused imports
* feat(openai/): add async `/image/variations` endpoint support
* feat(topaz/): support async `/image/variations` calls
* fix: test
* fix(utils.py): fix get_model_info_helper for no model info w/ provider config
handles situation where model info is not known but provider config exists
* test(test_router_fallbacks.py): mark flaky test
* fix: fix unused imports
* test: bump otel load test perf threshold - accounts for current load tests hitting same server
* feat(router.py): support passing model-specific messages in fallbacks
* docs(routing.md): separate router timeouts into separate doc
allow for 1 fallbacks doc (across proxy/router)
* docs(routing.md): cleanup router docs
* docs(reliability.md): cleanup docs
* docs(reliability.md): cleaned up fallback doc
just have 1 doc across sdk/proxy
simplifies docs
* docs(reliability.md): add setting model-specific fallback prompts
* fix: fix linting errors
* test: skip test causing openai rate limit errros
* test: fix test
* test: run vertex test first to catch error
* feat(bedrock/): add bedrock converse top k param
Closes https://github.com/BerriAI/litellm/issues/7087
* Fix bedrock empty content error (#7177)
* add resolver
* handle empty content on bedrock with default content
* use existing default message, tests
* Update tests/llm_translation/test_bedrock_completion.py
* fix tests
* Revert "add resolver"
This reverts commit c717e376ee.
* fallback to empty
---------
Co-authored-by: Krish Dholakia <krrishdholakia@gmail.com>
* fix(factory.py): handle empty content blocks in messages
Fixes https://github.com/BerriAI/litellm/issues/7169
* feat(router.py): add stripped model check to model fallback search
if model_name="openai/gpt-3.5-turbo" and fallback=[{"gpt-3.5-turbo"..}] the fallback should just work as expected
* fix: fix linting error
* fix(factory.py): fix linting error
* fix(factory.py): in base case still support skip empty text blocks
---------
Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
* fix(main.py): support passing max retries to azure/openai embedding integrations
Fixes https://github.com/BerriAI/litellm/issues/7003
* feat(team_endpoints.py): allow updating team model aliases
Closes https://github.com/BerriAI/litellm/issues/6956
* feat(router.py): allow specifying model id as fallback - skips any cooldown check
Allows a default model to be checked if all models in cooldown
s/o @micahjsmith
* docs(reliability.md): add fallback to specific model to docs
* fix(utils.py): new 'is_prompt_caching_valid_prompt' helper util
Allows user to identify if messages/tools have prompt caching
Related issue: https://github.com/BerriAI/litellm/issues/6784
* feat(router.py): store model id for prompt caching valid prompt
Allows routing to that model id on subsequent requests
* fix(router.py): only cache if prompt is valid prompt caching prompt
prevents storing unnecessary items in cache
* feat(router.py): support routing prompt caching enabled models to previous deployments
Closes https://github.com/BerriAI/litellm/issues/6784
* test: fix linting errors
* feat(databricks/): convert basemodel to dict and exclude none values
allow passing pydantic message to databricks
* fix(utils.py): ensure all chat completion messages are dict
* (feat) Track `custom_llm_provider` in LiteLLMSpendLogs (#7081)
* add custom_llm_provider to SpendLogsPayload
* add custom_llm_provider to SpendLogs
* add custom llm provider to SpendLogs payload
* test_spend_logs_payload
* Add MLflow to the side bar (#7031)
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>
* (bug fix) SpendLogs update DB catch all possible DB errors for retrying (#7082)
* catch DB_CONNECTION_ERROR_TYPES
* fix DB retry mechanism for SpendLog updates
* use DB_CONNECTION_ERROR_TYPES in auth checks
* fix exp back off for writing SpendLogs
* use _raise_failed_update_spend_exception to ensure errors print as NON blocking
* test_update_spend_logs_multiple_batches_with_failure
* (Feat) Add StructuredOutputs support for Fireworks.AI (#7085)
* fix model cost map fireworks ai "supports_response_schema": true,
* fix supports_response_schema
* fix map openai params fireworks ai
* test_map_response_format
* test_map_response_format
* added deepinfra/Meta-Llama-3.1-405B-Instruct (#7084)
* bump: version 1.53.9 → 1.54.0
* fix deepinfra
* litellm db fixes LiteLLM_UserTable (#7089)
* ci/cd queue new release
* fix llama-3.3-70b-versatile
* refactor - use consistent file naming convention `AI21/` -> `ai21` (#7090)
* fix refactor - use consistent file naming convention
* ci/cd run again
* fix naming structure
* fix use consistent naming (#7092)
---------
Signed-off-by: B-Step62 <yuki.watanabe@databricks.com>
Co-authored-by: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Co-authored-by: Yuki Watanabe <31463517+B-Step62@users.noreply.github.com>
Co-authored-by: ali sayyah <ali.sayyah2@gmail.com>
* fix(ollama.py): fix get model info request
Fixes https://github.com/BerriAI/litellm/issues/6703
* feat(anthropic/chat/transformation.py): support passing user id to anthropic via openai 'user' param
* docs(anthropic.md): document all supported openai params for anthropic
* test: fix tests
* fix: fix tests
* feat(jina_ai/): add rerank support
Closes https://github.com/BerriAI/litellm/issues/6691
* test: handle service unavailable error
* fix(handler.py): refactor together ai rerank call
* test: update test to handle overloaded error
* test: fix test
* Litellm router trace (#6742)
* feat(router.py): add trace_id to parent functions - allows tracking retry/fallbacks
* feat(router.py): log trace id across retry/fallback logic
allows grouping llm logs for the same request
* test: fix tests
* fix: fix test
* fix(transformation.py): only set non-none stop_sequences
* Litellm router disable fallbacks (#6743)
* bump: version 1.52.6 → 1.52.7
* feat(router.py): enable dynamically disabling fallbacks
Allows for enabling/disabling fallbacks per key
* feat(litellm_pre_call_utils.py): support setting 'disable_fallbacks' on litellm key
* test: fix test
* fix(exception_mapping_utils.py): map 'model is overloaded' to internal server error
* test: handle gemini error
* test: fix test
* fix: new run
* fix(deepseek/chat): convert content list to str
Fixes https://github.com/BerriAI/litellm/issues/6642
* test(test_deepseek_completion.py): implement base llm unit tests
increase robustness across providers
* fix(router.py): support content policy violation fallbacks with default fallbacks
* fix(opentelemetry.py): refactor to move otel imports behing flag
Fixes https://github.com/BerriAI/litellm/issues/6636
* fix(opentelemtry.py): close span on success completion
* fix(user_api_key_auth.py): allow user_role to default to none
* fix: mark flaky test
* fix(opentelemetry.py): move otelconfig.from_env to inside the init
prevent otel errors raised just by importing the litellm class
* fix(user_api_key_auth.py): fix auth error
* refactor(proxy_server.py): add debug logging around license check event (refactor position in startup_event logic)
* fix(proxy/_types.py): allow admin_allowed_routes to be any str
* fix(router.py): raise 400-status code error for no 'model_name' error on router
Fixes issue with status code when unknown model name passed with pattern matching enabled
* fix(converse_handler.py): add claude 3-5 haiku to bedrock converse models
* test: update testing to replace claude-instant-1.2
* fix(router.py): fix router.moderation calls
* test: update test to remove claude-instant-1
* fix(router.py): support model_list values in router.moderation
* test: fix test
* test: fix test
* feat(router.py): add check for max fallback depth
Prevent infinite loop for fallbacks
Closes https://github.com/BerriAI/litellm/issues/6498
* test: update test
* (fix) Prometheus - Log Postgres DB latency, status on prometheus (#6484)
* fix logging DB fails on prometheus
* unit testing log to otel wrapper
* unit testing for service logger + prometheus
* use LATENCY buckets for service logging
* fix service logging
* docs clarify vertex vs gemini
* (router_strategy/) ensure all async functions use async cache methods (#6489)
* fix router strat
* use async set / get cache in router_strategy
* add coverage for router strategy
* fix imports
* fix batch_get_cache
* use async methods for least busy
* fix least busy use async methods
* fix test_dual_cache_increment
* test async_get_available_deployment when routing_strategy="least-busy"
* (fix) proxy - fix when `STORE_MODEL_IN_DB` should be set (#6492)
* set store_model_in_db at the top
* correctly use store_model_in_db global
* (fix) `PrometheusServicesLogger` `_get_metric` should return metric in Registry (#6486)
* fix logging DB fails on prometheus
* unit testing log to otel wrapper
* unit testing for service logger + prometheus
* use LATENCY buckets for service logging
* fix service logging
* fix _get_metric in prom services logger
* add clear doc string
* unit testing for prom service logger
* bump: version 1.51.0 → 1.51.1
* Add `azure/gpt-4o-mini-2024-07-18` to model_prices_and_context_window.json (#6477)
* Update utils.py (#6468)
Fixed missing keys
* (perf) Litellm redis router fix - ~100ms improvement (#6483)
* docs(exception_mapping.md): add missing exception types
Fixes https://github.com/Aider-AI/aider/issues/2120#issuecomment-2438971183
* fix(main.py): register custom model pricing with specific key
Ensure custom model pricing is registered to the specific model+provider key combination
* test: make testing more robust for custom pricing
* fix(redis_cache.py): instrument otel logging for sync redis calls
ensures complete coverage for all redis cache calls
* refactor: pass parent_otel_span for redis caching calls in router
allows for more observability into what calls are causing latency issues
* test: update tests with new params
* refactor: ensure e2e otel tracing for router
* refactor(router.py): add more otel tracing acrosss router
catch all latency issues for router requests
* fix: fix linting error
* fix(router.py): fix linting error
* fix: fix test
* test: fix tests
* fix(dual_cache.py): pass ttl to redis cache
* fix: fix param
* perf(cooldown_cache.py): improve cooldown cache, to store cache results in memory for 5s, prevents redis call from being made on each request
reduces 100ms latency per call with caching enabled on router
* fix: fix test
* fix(cooldown_cache.py): handle if a result is None
* fix(cooldown_cache.py): add debug statements
* refactor(dual_cache.py): move to using an in-memory check for batch get cache, to prevent redis from being hit for every call
* fix(cooldown_cache.py): fix linting erropr
* build: merge main
---------
Co-authored-by: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Co-authored-by: Xingyao Wang <xingyao@all-hands.dev>
Co-authored-by: vibhanshu-ob <115142120+vibhanshu-ob@users.noreply.github.com>
* docs(exception_mapping.md): add missing exception types
Fixes https://github.com/Aider-AI/aider/issues/2120#issuecomment-2438971183
* fix(main.py): register custom model pricing with specific key
Ensure custom model pricing is registered to the specific model+provider key combination
* test: make testing more robust for custom pricing
* fix(redis_cache.py): instrument otel logging for sync redis calls
ensures complete coverage for all redis cache calls
