* 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
* test: unshadow the module handles the F811 sweep left behind, and pin the two live tests that went red with it
The F811 sweep in #37878 removed the fixture-local `import litellm` from four
conftests, but the bare `import litellm.proxy.proxy_server` a few lines below
still binds `litellm` as a function local, so `importlib.reload(litellm)` runs
before the name is assigned and every test in those directories errors at
setup. The `hasattr` guard on the line above already proves the module is
loaded, so the import only ever bound the name. Drop it, and enable F823 in
ruff-tests.toml, which flags all four sites at the failing line and would have
blocked the sweep
The same sweep renamed the `check_non_streaming_response` parameter but left
one read of `completion`, which now resolves to `litellm.completion`, and
removed an import whose side effect was the only thing making
`litellm.proxy.proxy_server` reachable in the moderation hook test. That test
already takes `monkeypatch`, so patch the router through it and stop leaking
the router into later tests
`test_content_policy_exception_openai` passed vacuously until #37887 turned it
into a real `pytest.raises`, and OpenAI no longer rejects a lyrics prompt with
a content policy error. Inject an AsyncOpenAI client whose transport answers
with OpenAI's own `content_policy_violation` rejection so the mapping to
ContentPolicyViolationError is exercised every run
`test_async_create_batch` hit a 409 cancelling a batch OpenAI had already
marked failed. The cancel step tolerated a completed batch but not a failed
one. Fold both guards into one helper that tolerates a failed batch only when
OpenAI's recorded error is the org's enqueued token limit, and prints the
batch's errors so the reason is in the log either way
* test: close the injected AsyncOpenAI client after the content policy test
* chore(lint): ratchet TQ005 down by the global mutation this branch cleared
* chore(lint): ratchet TQ005 to 2660 on the merged tree
* chore(lint): ratchet TQ005 to 2561 on the merged tree
* chore(lint): ratchet TQ005 to 2548 on the merged tree
A name bound twice keeps only the second binding. In `tests/` that is nearly
always a repeated import, harmless but misleading, and the same rule is what
catches the cases that are not harmless: a local that shadows an import the
module still calls, and a second `def test_x` that quietly replaces the first.
311 of the 344 sites were repeated imports and came out with ruff's own fix.
The remaining 33 needed a decision. Four modules imported a name they never
used because a local definition below already shadowed it. Two comprehensions
bound `call` over `unittest.mock.call`, which those modules import and use.
One test rebound the two module handles its nested reload closure had captured.
One class attribute shadowed an unused `status` import.
The load-test fixtures move to a conftest, which is how pytest is meant to share
them, so the test module no longer imports three fixture names it never calls.
The nine `prisma_client` parameters keep a narrow `noqa`: pytest resolves that
fixture by name before the body runs, so the parameter never shadows anything.
The earlier sweep only caught the conformance suite in tests/llm_translation.
Groq retired llama-3.1-8b-instant alongside llama-3.3-70b-versatile, and four
tests under tests/local_testing still call them for real, so litellm_router_testing
and both local_testing shards 404 with model_not_found.
Only the sites that leave the process move. The chunk fixtures in
test_stream_chunk_builder, and the cost and routing tests that never open a
socket, keep the old ids because the string is data there, not a request.
* fix(tests): replace shut-down gpt-4o-audio-preview with gpt-audio-1.5
OpenAI shut down gpt-4o-audio-preview on 2026-05-07, so the live audio
calls in test_stream_chunk_builder_openai_audio_output_usage and
test_standard_logging_payload_audio now hard-fail with a model-not-found
error on every PR. The error was not "openai-internal", so the except
block swallowed it and execution fell through to an unbound
completion/response (UnboundLocalError).
Switch both tests to gpt-audio-1.5, OpenAI's recommended successor
(GA, not deprecated, already present in the litellm cost map so the
response_cost assertion still resolves). Also broaden the except to
skip with the real error in the reason instead of crashing, so a
transient upstream blip can't reintroduce the UnboundLocalError.
* fix(tests): narrow audio-test skip to model-not-found, re-raise the rest
Address review feedback: an unconditional skip on any exception would
silently mask a litellm-internal regression in the audio path (broken
param transformation, serialization, bad header) instead of failing CI.
Skip only on the upstream-unavailable class (model_not_found / "does not
exist" / openai-internal) and re-raise everything else, so genuine
regressions still fail loudly. The UnboundLocalError is still fixed
because the handler either skips or raises - it never falls through.
* fix(tests): add budget_exceeded to expected Interaction status enum
Staging added budget_exceeded to the Interaction OpenAPI status enum; the staging merge into this branch picked up the spec change but not the matching test update, so test_status_enum_values failed in CI. Align the test's expected list (exact-match by design) with the live spec.
* fix(tests): mock HTTP fetch in test_img_url_token_counter
The test parameterized a live third-party image URL (blog.purpureus.net) which now 404s, causing get_image_dimensions to fall through to its base64 decode path and crash with 'not enough values to unpack' on every PR run. Mock safe_get with a tiny 1x1 PNG so the URL branch is still exercised without any network dependency.
* fix(tests): swap gpt-4o-audio-preview to gpt-audio-1.5 in test_gpt4o_audio
OpenAI shut down gpt-4o-audio-preview on 2026-05-07, so both live tests in test_gpt4o_audio.py (test_audio_output_from_model and test_audio_input_to_model) hard-fail model_not_found on every PR. Swap the hardcoded model to OpenAI's successor gpt-audio-1.5 (same chat-completions audio surface; already in the litellm cost map). Mirror the narrowed-skip pattern from the prior audio fixes: skip on model_not_found / does-not-exist / openai-internal, re-raise everything else so genuine litellm regressions still fail CI loudly.
The test calls OpenAI's gpt-4o-audio-preview model which sometimes
doesn't return usage data in the streaming response. Fixed by:
- Adding @pytest.mark.flaky(retries=5, delay=2) for retry handling
- Fixing usage_obj loop to check chunk.usage is not None
- Skipping gracefully when OpenAI doesn't return usage data
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Replace ModelResponse(stream=True) with ModelResponseStream in
test_unit_test_custom_stream_wrapper_repeating_chunk — stream=True
stores delta as a plain dict causing AttributeError in CustomStreamWrapper
- Accept MidStreamFallbackError alongside InternalServerError in the
repeating-chunk safety check assertion
- Add @pytest.mark.flaky(retries=3) to the live OpenAI audio output
usage test
ModelResponse.choices was typed as List[Union[Choices, StreamingChoices]] which
caused Pydantic serialization warnings and false linting errors. Now that
ModelResponseStream exists for streaming, narrow ModelResponse.choices to
List[Choices] and migrate all ModelResponse(stream=True) call sites to use
ModelResponseStream() instead.
* fix(vertex_ai/gemini/transformation.py): handle 'http://' image urls
* test: add base test for `http:` url's
* fix(factory.py/get_image_details): follow redirects
allows http calls to work
* fix(codestral/): fix stream chunk parsing on last chunk of stream
* Azure ad token provider (#6917)
* Update azure.py
Added optional parameter azure ad token provider
* Added parameter to main.py
* Found token provider arg location
* Fixed embeddings
* Fixed ad token provider
---------
Co-authored-by: Krish Dholakia <krrishdholakia@gmail.com>
* fix: fix linting errors
* fix(main.py): leave out o1 route for azure ad token provider, for now
get v0 out for sync azure gpt route to begin with
* test: skip http:// test for fireworks ai
model does not support it
* refactor: cleanup dead code
* fix: revert http:// url passthrough for gemini
google ai studio raises errors
* test: fix test
---------
Co-authored-by: bahtman <anton@baht.dk>
* ui 1 - show correct msg on no logs
* fix dup country col
* backend - allow filtering by team_id and api_key
* fix ui_view_spend_logs
* ui update query params
* working team id and key hash filters
* fix filter ref - don't hold on them as they are
* fix _model_custom_llm_provider_matches_wildcard_pattern
* fix test test_stream_chunk_builder_openai_audio_output_usage - use direct dict comparison
* refactor(factory.py): refactor async bedrock message transformation to use async get request for image url conversion
improve latency of bedrock call
* test(test_bedrock_completion.py): add unit testing to ensure async image url get called for async bedrock call
* refactor(factory.py): refactor bedrock translation to use BedrockImageProcessor
reduces duplicate code
* fix(factory.py): fix bug not allowing pdf's to be processed
* fix(factory.py): fix bedrock converse document understanding with image url
* docs(bedrock.md): clarify all bedrock document types are supported
* refactor: cleanup redundant test + unused imports
* perf: improve perf with reusable clients
* test: fix test
* fix(streaming_chunk_builder_utils.py): add test for groq tool calling + streaming + combine chunks
Addresses https://github.com/BerriAI/litellm/issues/7621
* fix(streaming_utils.py): fix modelresponseiterator for openai like chunk parser
ensures chunk parser uses the correct tool call id when translating the chunk
Fixes https://github.com/BerriAI/litellm/issues/7621
* build(model_hub.tsx): display cost pricing on model hub
* build(model_hub.tsx): show cost per token pricing + complete model information
* fix(types/utils.py): fix usage object handling
* fix(invoke_handler.py): fix mock response iterator to handle tool calling
returns tool call if returned by model response
* fix(prometheus.py): add new 'tokens_by_tag' metric on prometheus
allows tracking 'token usage' by task
* feat(prometheus.py): add input + output token tracking by tag
* feat(prometheus.py): add tag based deployment failure tracking
allows admin to track failure by use-case
* fix(cost_calculator.py): move to using `.get_model_info()` for cost per token calculations
ensures cost tracking is reliable - handles edge cases of parsing model cost map
* build(model_prices_and_context_window.json): add 'supports_response_schema' for select tgai models
Fixes https://github.com/BerriAI/litellm/pull/7037#discussion_r1872157329
* build(model_prices_and_context_window.json): remove 'pdf input' and 'vision' support from nova micro in model map
Bedrock docs indicate no support for micro - https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html
* fix(converse_transformation.py): support amazon nova tool use
* fix(opentelemetry): Add missing LLM request type attribute to spans (#7041)
* feat(opentelemetry): add LLM request type attribute to spans
* lint
* fix: curl usage (#7038)
curl -d, --data <data> is lowercase d
curl -D, --dump-header <filename> is uppercase D
references:
https://curl.se/docs/manpage.html#-dhttps://curl.se/docs/manpage.html#-D
* fix(spend_tracking.py): handle empty 'id' in model response - when creating spend log
Fixes https://github.com/BerriAI/litellm/issues/7023
* fix(streaming_chunk_builder.py): handle initial id being empty string
Fixes https://github.com/BerriAI/litellm/issues/7023
* fix(anthropic_passthrough_logging_handler.py): add end user cost tracking for anthropic pass through endpoint
* docs(pass_through/): refactor docs location + add table on supported features for pass through endpoints
* feat(anthropic_passthrough_logging_handler.py): support end user cost tracking via anthropic sdk
* docs(anthropic_completion.md): add docs on passing end user param for cost tracking on anthropic sdk
* fix(litellm_logging.py): use standard logging payload if present in kwargs
prevent datadog logging error for pass through endpoints
* docs(bedrock.md): add rerank api usage example to docs
* bugfix/change dummy tool name format (#7053)
* fix viewing keys (#7042)
* ui new build
* build(model_prices_and_context_window.json): add bedrock region models to model cost map (#7044)
* bye (#6982)
* (fix) litellm router.aspeech (#6962)
* doc Migrating Databases
* fix aspeech on router
* test_audio_speech_router
* test_audio_speech_router
* docs show supported providers on batches api doc
* change dummy tool name format
---------
Co-authored-by: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Co-authored-by: Krish Dholakia <krrishdholakia@gmail.com>
Co-authored-by: yujonglee <yujonglee.dev@gmail.com>
* fix: fix linting errors
* test: update test
* fix(litellm_logging.py): fix pass through check
* fix(test_otel_logging.py): fix test
* fix(cost_calculator.py): update handling for cost per second
* fix(cost_calculator.py): fix cost check
* test: fix test
* (fix) adding public routes when using custom header (#7045)
* get_api_key_from_custom_header
* add test_get_api_key_from_custom_header
* fix testing use 1 file for test user api key auth
* fix test user api key auth
* test_custom_api_key_header_name
* build: update ui build
---------
Co-authored-by: Doron Kopit <83537683+doronkopit5@users.noreply.github.com>
Co-authored-by: lloydchang <lloydchang@gmail.com>
Co-authored-by: hgulersen <haymigulersen@gmail.com>
Co-authored-by: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Co-authored-by: yujonglee <yujonglee.dev@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
* refactor(main.py): streaming_chunk_builder
use <100 lines of code
refactor each component into a separate function - easier to maintain + test
* fix(utils.py): handle choices being None
openai pydantic schema updated
* fix(main.py): fix linting error
* feat(streaming_chunk_builder_utils.py): update stream chunk builder to support rebuilding audio chunks from openai
* test(test_custom_callback_input.py): test message redaction works for audio output
* fix(streaming_chunk_builder_utils.py): return anthropic token usage info directly
* fix(stream_chunk_builder_utils.py): run validation check before entering chunk processor
* fix(main.py): fix import
