Skip to content

fix(openai): Simplify Completions input handling #5690

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 9 commits into from
Mar 26, 2026
Merged

fix(openai): Simplify Completions input handling #5690

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 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
0e7078d
fix(openai): Simplify Completions input handling
alexander-alderman-webb Mar 18, 2026
4de8cdb
Merge branch 'master' into webb/openai/remove-common-input-data
alexander-alderman-webb Mar 25, 2026
c11c0f6
Merge branch 'webb/openai/remove-common-input-data' into webb/openai/…
alexander-alderman-webb Mar 25, 2026
fd908c5
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
20 changes: 8 additions & 12 deletions sentry_sdk/integrations/openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -356,36 +356,32 @@ def _set_completions_api_input_data(
)

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

if (
not should_send_default_pii()
Expand Down
100 changes: 100 additions & 0 deletions tests/integrations/openai/test_openai.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -151,6 +151,11 @@ def test_nonstreaming_chat_completion_no_prompts(
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "hello"},
],
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)
.choices[0]
.message.content
Expand All @@ -163,6 +168,13 @@ def test_nonstreaming_chat_completion_no_prompts(
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

assert SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS not in span["data"]
assert SPANDATA.GEN_AI_REQUEST_MESSAGES not in span["data"]
assert SPANDATA.GEN_AI_RESPONSE_TEXT not in span["data"]
Expand Down Expand Up @@ -231,6 +243,11 @@ def test_nonstreaming_chat_completion(sentry_init, capture_events, messages, req
client.chat.completions.create(
model="some-model",
messages=messages,
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)
.choices[0]
.message.content
Expand All @@ -243,6 +260,13 @@ def test_nonstreaming_chat_completion(sentry_init, capture_events, messages, req
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

param_id = request.node.callspec.id
if "blocks" in param_id:
assert json.loads(span["data"][SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS]) == [
Expand Down Expand Up @@ -300,6 +324,11 @@ async def test_nonstreaming_chat_completion_async_no_prompts(
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "hello"},
],
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)
response = response.choices[0].message.content

Expand All @@ -310,6 +339,13 @@ async def test_nonstreaming_chat_completion_async_no_prompts(
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

assert SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS not in span["data"]
assert SPANDATA.GEN_AI_REQUEST_MESSAGES not in span["data"]
assert SPANDATA.GEN_AI_RESPONSE_TEXT not in span["data"]
Expand Down Expand Up @@ -380,6 +416,11 @@ async def test_nonstreaming_chat_completion_async(
response = await client.chat.completions.create(
model="some-model",
messages=messages,
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)
response = response.choices[0].message.content

Expand All @@ -390,6 +431,13 @@ async def test_nonstreaming_chat_completion_async(
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

param_id = request.node.callspec.id
if "blocks" in param_id:
assert json.loads(span["data"][SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS]) == [
Expand Down Expand Up @@ -498,6 +546,11 @@ def test_streaming_chat_completion_no_prompts(
{"role": "user", "content": "hello"},
],
stream=True,
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)
response_string = "".join(
map(lambda x: x.choices[0].delta.content, response_stream)
Expand All @@ -508,6 +561,14 @@ def test_streaming_chat_completion_no_prompts(
span = tx["spans"][0]
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
assert span["data"][SPANDATA.GEN_AI_RESPONSE_STREAMING] is True

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

assert span["data"][SPANDATA.GEN_AI_RESPONSE_MODEL] == "model-id"

Expand Down Expand Up @@ -626,6 +687,11 @@ def test_streaming_chat_completion(sentry_init, capture_events, messages, reques
model="some-model",
messages=messages,
stream=True,
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)
response_string = "".join(
map(lambda x: x.choices[0].delta.content, response_stream)
Expand All @@ -636,6 +702,14 @@ def test_streaming_chat_completion(sentry_init, capture_events, messages, reques
span = tx["spans"][0]
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
assert span["data"][SPANDATA.GEN_AI_RESPONSE_STREAMING] is True

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

param_id = request.node.callspec.id
if "blocks" in param_id:
Expand Down Expand Up @@ -753,6 +827,11 @@ async def test_streaming_chat_completion_async_no_prompts(
{"role": "user", "content": "hello"},
],
stream=True,
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)

response_string = ""
Expand All @@ -765,6 +844,14 @@ async def test_streaming_chat_completion_async_no_prompts(
span = tx["spans"][0]
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
assert span["data"][SPANDATA.GEN_AI_RESPONSE_STREAMING] is True

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

assert span["data"][SPANDATA.GEN_AI_RESPONSE_MODEL] == "model-id"

Expand Down Expand Up @@ -891,6 +978,11 @@ async def test_streaming_chat_completion_async(
model="some-model",
messages=messages,
stream=True,
max_tokens=100,
presence_penalty=0.1,
frequency_penalty=0.2,
temperature=0.7,
top_p=0.9,
)

response_string = ""
Expand All @@ -903,6 +995,14 @@ async def test_streaming_chat_completion_async(
span = tx["spans"][0]
assert span["op"] == "gen_ai.chat"
assert span["data"][SPANDATA.GEN_AI_SYSTEM] == "openai"
assert span["data"][SPANDATA.GEN_AI_RESPONSE_STREAMING] is True

assert span["data"][SPANDATA.GEN_AI_REQUEST_MODEL] == "some-model"
assert span["data"][SPANDATA.GEN_AI_REQUEST_MAX_TOKENS] == 100
assert span["data"][SPANDATA.GEN_AI_REQUEST_PRESENCE_PENALTY] == 0.1
assert span["data"][SPANDATA.GEN_AI_REQUEST_FREQUENCY_PENALTY] == 0.2
assert span["data"][SPANDATA.GEN_AI_REQUEST_TEMPERATURE] == 0.7
assert span["data"][SPANDATA.GEN_AI_REQUEST_TOP_P] == 0.9

assert span["data"][SPANDATA.GEN_AI_RESPONSE_MODEL] == "model-id"

Expand Down
Loading