Skip to content

fix(openai): Use max_output_tokens for Responses API #5693

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
alexander-alderman-webb merged 10 commits into from
Mar 26, 2026
+14 −2
Merged

fix(openai): Use max_output_tokens for Responses API #5693

Show file tree
Hide file tree
Changes from 9 commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
4917b6a
ref(openai): Move input handling code into API-specific functions
alexander-alderman-webb Mar 18, 2026
4261546
keep set_data_normalized
alexander-alderman-webb Mar 18, 2026
1f6149e
.
alexander-alderman-webb Mar 18, 2026
36842e0
.
alexander-alderman-webb Mar 18, 2026
532fa65
.
alexander-alderman-webb Mar 18, 2026
9fb4bff
fix(openai): Use max_output_tokens for Responses API
alexander-alderman-webb Mar 18, 2026
9935178
.
alexander-alderman-webb Mar 18, 2026
4de8cdb
Merge branch 'master' into webb/openai/remove-common-input-data
alexander-alderman-webb Mar 25, 2026
cef5444
Merge branch 'webb/openai/remove-common-input-data' into webb/openai/…
alexander-alderman-webb Mar 25, 2026
40acdf1
merge master
alexander-alderman-webb Mar 26, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
155 changes: 114 additions & 41 deletions sentry_sdk/integrations/openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -219,48 +219,54 @@ def _calculate_token_usage(
)


def _commmon_set_input_data(
def _set_responses_api_input_data(
span: "Span",
kwargs: "dict[str, Any]",
integration: "OpenAIIntegration",
) -> None:
# Input attributes: Common
set_data_normalized(span, SPANDATA.GEN_AI_SYSTEM, "openai")

# Input attributes: Optional
kwargs_keys_to_attributes = {
"model": SPANDATA.GEN_AI_REQUEST_MODEL,
"stream": SPANDATA.GEN_AI_RESPONSE_STREAMING,
"max_tokens": SPANDATA.GEN_AI_REQUEST_MAX_TOKENS,
"presence_penalty": SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY,
"frequency_penalty": SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY,
"temperature": SPANDATA.GEN_AI_REQUEST_TEMPERATURE,
"top_p": SPANDATA.GEN_AI_REQUEST_TOP_P,
}
for key, attribute in kwargs_keys_to_attributes.items():
value = kwargs.get(key)

if value is not None and _is_given(value):
set_data_normalized(span, attribute, value)

# Input attributes: Tools
explicit_instructions: "Union[Optional[str], Omit]" = kwargs.get("instructions")
messages: "Optional[Union[str, ResponseInputParam]]" = kwargs.get("input")

tools = kwargs.get("tools")
if tools is not None and _is_given(tools) and len(tools) > 0:
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_AVAILABLE_TOOLS, safe_serialize(tools)
)

model = kwargs.get("model")
if model is not None and _is_given(model):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MODEL, model)

def _set_responses_api_input_data(
span: "Span",
kwargs: "dict[str, Any]",
integration: "OpenAIIntegration",
) -> None:
explicit_instructions: "Union[Optional[str], Omit]" = kwargs.get("instructions")
messages: "Optional[Union[str, ResponseInputParam]]" = kwargs.get("input")
stream = kwargs.get("stream")
if stream is not None and _is_given(stream):
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_STREAMING, stream)

max_tokens = kwargs.get("max_output_tokens")
if max_tokens is not None and _is_given(max_tokens):
span.set_data(SPANDATA.GEN_AI_REQUEST_MAX_TOKENS, max_tokens)

presence_penalty = kwargs.get("presence_penalty")
if presence_penalty is not None and _is_given(presence_penalty):
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY, presence_penalty
)

frequency_penalty = kwargs.get("frequency_penalty")
if frequency_penalty is not None and _is_given(frequency_penalty):
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY, frequency_penalty
)

temperature = kwargs.get("temperature")
if temperature is not None and _is_given(temperature):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TEMPERATURE, temperature)

top_p = kwargs.get("top_p")
if top_p is not None and _is_given(top_p):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TOP_P, top_p)

if not should_send_default_pii() or not integration.include_prompts:
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "responses")
_commmon_set_input_data(span, kwargs)
return

if (
Expand All @@ -281,12 +287,10 @@ def _set_responses_api_input_data(
)

set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "responses")
_commmon_set_input_data(span, kwargs)
return

if messages is None:
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "responses")
_commmon_set_input_data(span, kwargs)
return

instructions_text_parts: "list[TextPart]" = []
Expand Down Expand Up @@ -319,7 +323,6 @@ def _set_responses_api_input_data(
)

set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "responses")
_commmon_set_input_data(span, kwargs)
return

non_system_messages = [
Expand All @@ -335,7 +338,6 @@ def _set_responses_api_input_data(
)

set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "responses")
_commmon_set_input_data(span, kwargs)


def _set_completions_api_input_data(
Expand All @@ -347,13 +349,50 @@ def _set_completions_api_input_data(
"messages"
)

tools = kwargs.get("tools")
if tools is not None and _is_given(tools) and len(tools) > 0:
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_AVAILABLE_TOOLS, safe_serialize(tools)
)

model = kwargs.get("model")
if model is not None and _is_given(model):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MODEL, model)

stream = kwargs.get("stream")
if stream is not None and _is_given(stream):
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_STREAMING, stream)

max_tokens = kwargs.get("max_tokens")
if max_tokens is not None and _is_given(max_tokens):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MAX_TOKENS, max_tokens)

presence_penalty = kwargs.get("presence_penalty")
if presence_penalty is not None and _is_given(presence_penalty):
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY, presence_penalty
)

frequency_penalty = kwargs.get("frequency_penalty")
if frequency_penalty is not None and _is_given(frequency_penalty):
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY, frequency_penalty
)

temperature = kwargs.get("temperature")
if temperature is not None and _is_given(temperature):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TEMPERATURE, temperature)

top_p = kwargs.get("top_p")
if top_p is not None and _is_given(top_p):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TOP_P, top_p)

if (
not should_send_default_pii()
or not integration.include_prompts
or messages is None
):
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "chat")
_commmon_set_input_data(span, kwargs)
return

if isinstance(messages, str):
Expand All @@ -365,13 +404,11 @@ def _set_completions_api_input_data(
span, SPANDATA.GEN_AI_REQUEST_MESSAGES, messages_data, unpack=False
)
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "chat")
_commmon_set_input_data(span, kwargs)
return

# dict special case following https://github.com/openai/openai-python/blob/3e0c05b84a2056870abf3bd6a5e7849020209cc3/src/openai/_utils/_transform.py#L194-L197
if not isinstance(messages, Iterable) or isinstance(messages, dict):
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "chat")
_commmon_set_input_data(span, kwargs)
return

messages = list(messages)
Expand Down Expand Up @@ -399,7 +436,6 @@ def _set_completions_api_input_data(
)

set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "chat")
_commmon_set_input_data(span, kwargs)


def _set_embeddings_input_data(
Expand All @@ -411,19 +447,55 @@ def _set_embeddings_input_data(
"input"
)

tools = kwargs.get("tools")
if tools is not None and _is_given(tools) and len(tools) > 0:
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_AVAILABLE_TOOLS, safe_serialize(tools)
)

model = kwargs.get("model")
if model is not None and _is_given(model):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MODEL, model)

stream = kwargs.get("stream")
if stream is not None and _is_given(stream):
set_data_normalized(span, SPANDATA.GEN_AI_RESPONSE_STREAMING, stream)

max_tokens = kwargs.get("max_tokens")
if max_tokens is not None and _is_given(max_tokens):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_MAX_TOKENS, max_tokens)

presence_penalty = kwargs.get("presence_penalty")
if presence_penalty is not None and _is_given(presence_penalty):
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY, presence_penalty
)

frequency_penalty = kwargs.get("frequency_penalty")
if frequency_penalty is not None and _is_given(frequency_penalty):
set_data_normalized(
span, SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY, frequency_penalty
)

temperature = kwargs.get("temperature")
if temperature is not None and _is_given(temperature):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TEMPERATURE, temperature)

top_p = kwargs.get("top_p")
if top_p is not None and _is_given(top_p):
set_data_normalized(span, SPANDATA.GEN_AI_REQUEST_TOP_P, top_p)

if (
not should_send_default_pii()
or not integration.include_prompts
or messages is None
):
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "embeddings")
_commmon_set_input_data(span, kwargs)

return

if isinstance(messages, str):
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "embeddings")
_commmon_set_input_data(span, kwargs)

normalized_messages = normalize_message_roles([messages]) # type: ignore
scope = sentry_sdk.get_current_scope()
Expand All @@ -440,7 +512,6 @@ def _set_embeddings_input_data(
# dict special case following https://github.com/openai/openai-python/blob/3e0c05b84a2056870abf3bd6a5e7849020209cc3/src/openai/_utils/_transform.py#L194-L197
if not isinstance(messages, Iterable) or isinstance(messages, dict):
set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "embeddings")
_commmon_set_input_data(span, kwargs)
return

messages = list(messages)
Expand All @@ -458,7 +529,6 @@ def _set_embeddings_input_data(
)

set_data_normalized(span, SPANDATA.GEN_AI_OPERATION_NAME, "embeddings")
_commmon_set_input_data(span, kwargs)


def _set_common_output_data(
Expand Down Expand Up @@ -551,6 +621,7 @@ def _new_chat_completion_common(f: "Any", *args: "Any", **kwargs: "Any") -> "Any
)
span.__enter__()

span.set_data(SPANDATA.GEN_AI_SYSTEM, "openai")
_set_completions_api_input_data(span, kwargs, integration)

start_time = time.perf_counter()
Expand Down Expand Up @@ -945,6 +1016,7 @@ def _new_embeddings_create_common(f: "Any", *args: "Any", **kwargs: "Any") -> "A
name=f"embeddings {model}",
origin=OpenAIIntegration.origin,
) as span:
span.set_data(SPANDATA.GEN_AI_SYSTEM, "openai")
_set_embeddings_input_data(span, kwargs, integration)

response = yield f, args, kwargs
Expand Down Expand Up @@ -1036,6 +1108,7 @@ def _new_responses_create_common(f: "Any", *args: "Any", **kwargs: "Any") -> "An
)
span.__enter__()

span.set_data(SPANDATA.GEN_AI_SYSTEM, "openai")
_set_responses_api_input_data(span, kwargs, integration)

start_time = time.perf_counter()
Expand Down
Loading
Loading