Skip to content

Commit 79d7785

Browse files
apcha-oaiapcha-oai
authored
Merge 6aae2f5 into 03b3ec4 
2 parents 03b3ec4 + 6aae2f5 commit 79d7785

165 files changed

Lines changed: 5954 additions & 2681 deletions

File tree

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎.castiron.stats.yml‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
schema_version: 1
2-
generation_id: e1943695-9da0-4ac1-ae5a-cc763f5d90be
2+
generation_id: 6f16b5a1-6b31-43ea-a1dd-a1a9fc37821e
33
openapi_spec_hash: 1faf0319c407c6534ef7c381614afbb6
44
openapi_transformed_spec_hash: 53eff50caa9d18046e4ff0615bc7512d
55
config_hash: 4617b0962f16d328804312ac81aac630
6-
codegen_sha: b97c76f867d26f99d0de01d6283fa6d511044141
6+
codegen_sha: 20f59b5cced5772bda4c6d6e3e19726d3805e4a1

‎README.md‎

Lines changed: 21 additions & 69 deletions
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,7 @@
55

66
The OpenAI Python library provides convenient access to the OpenAI REST API from any Python 3.10+
77
application. The library includes type definitions for all request params and response fields,
8-
and offers both synchronous and asynchronous clients powered by [httpx](https://github.com/encode/httpx).
8+
and offers both synchronous and asynchronous clients powered by [HTTPX2](https://httpx2.pydantic.dev/).
99

1010
It is generated from our [OpenAPI specification](https://github.com/openai/openai-openapi) with [Stainless](https://stainlessapi.com/).
1111

@@ -244,9 +244,7 @@ Functionality between the synchronous and asynchronous clients is otherwise iden
244244

245245
### With aiohttp
246246

247-
By default, the async client uses `httpx` for HTTP requests. However, for improved concurrency performance you may also use `aiohttp` as the HTTP backend.
248-
249-
The `aiohttp` backend requires Python 3.10 or later.
247+
By default, the async client uses HTTPX2. For improved concurrency performance, you may also use `aiohttp` as the HTTPX2 transport.
250248

251249
You can enable this by installing `aiohttp`:
252250

@@ -283,32 +281,9 @@ async def main() -> None:
283281
asyncio.run(main())
284282
```
285283

286-
### Experimental HTTPX2 support
287-
288-
To opt in to experimental HTTPX2 support, install the optional extra on Python 3.10 or later:
289-
290-
```sh
291-
pip install 'openai[httpx2]'
292-
```
293-
294-
```python
295-
from openai import OpenAI, AsyncOpenAI, DefaultHttpx2Client, DefaultAsyncHttpx2Client
296-
297-
client = OpenAI(http_client=DefaultHttpx2Client())
298-
async_client = AsyncOpenAI(http_client=DefaultAsyncHttpx2Client())
299-
```
300-
301-
See [`examples/httpx2_client.py`](examples/httpx2_client.py) for a minimal runnable example.
284+
### HTTPX2 migration
302285

303-
The module-level client can be configured in the same way:
304-
305-
```python
306-
import openai
307-
308-
openai.http_client = openai.DefaultHttpx2Client()
309-
```
310-
311-
Parsed API models are unchanged, but requests, raw and streaming responses, and transport-level exceptions may be HTTPX2 objects at runtime. Code that catches HTTPX exceptions or relies on HTTPX-specific mocks, transports, authentication, hooks, or instrumentation may need to be updated. Transport-facing type annotations may still describe HTTPX.
286+
HTTPX2 is the default HTTP client. If you configure a custom HTTP client, transport, timeout, authentication handler, event hook, or request mock, see the [HTTPX2 migration guide](httpx2.md).
312287

313288
## Streaming responses
314289

@@ -636,7 +611,7 @@ try:
636611
)
637612
except openai.APIConnectionError as e:
638613
print("The server could not be reached")
639-
print(e.__cause__) # an underlying Exception, likely raised within httpx.
614+
print(e.__cause__) # an underlying Exception, likely raised within HTTPX2.
640615
except openai.RateLimitError as e:
641616
print("A 429 status code was received; we should back off a bit.")
642617
except openai.APIStatusError as e:
@@ -723,9 +698,10 @@ client.with_options(max_retries=5).chat.completions.create(
723698
## Timeouts
724699

725700
By default requests time out after 10 minutes. You can configure this with a `timeout` option,
726-
which accepts a float or an [`httpx.Timeout`](https://www.python-httpx.org/advanced/timeouts/#fine-tuning-the-configuration) object:
701+
which accepts a float or an [`httpx2.Timeout`](https://httpx2.pydantic.dev/) object:
727702

728703
```python
704+
import httpx2
729705
from openai import OpenAI
730706

731707
# Configure the default for all requests:
@@ -736,7 +712,7 @@ client = OpenAI(
736712

737713
# More granular control:
738714
client = OpenAI(
739-
timeout=httpx.Timeout(60.0, read=5.0, write=10.0, connect=2.0),
715+
timeout=httpx2.Timeout(60.0, read=5.0, write=10.0, connect=2.0),
740716
)
741717

742718
# Override per-request:
@@ -849,11 +825,11 @@ To make requests to undocumented endpoints, you can make requests using `client.
849825
http verbs. Options on the client will be respected (such as retries) when making this request.
850826

851827
```py
852-
import httpx
828+
import httpx2
853829

854830
response = client.post(
855831
"/foo",
856-
cast_to=httpx.Response,
832+
cast_to=httpx2.Response,
857833
body={"my_param": True},
858834
)
859835

@@ -873,30 +849,26 @@ can also get all the extra fields on the Pydantic model as a dict with
873849

874850
### Configuring the HTTP client
875851

876-
You can directly override the [httpx client](https://www.python-httpx.org/api/#client) to customize it for your use case, including:
877-
878-
- Support for [proxies](https://www.python-httpx.org/advanced/proxies/)
879-
- Custom [transports](https://www.python-httpx.org/advanced/transports/)
880-
- Additional [advanced](https://www.python-httpx.org/advanced/clients/) functionality
852+
You can override the [HTTPX2 client](https://httpx2.pydantic.dev/) to customize proxies, transports, authentication, event hooks, and other advanced HTTP behavior. See the [HTTPX2 migration guide](httpx2.md) when updating an existing custom client.
881853

882854
```python
883-
import httpx
884-
from openai import OpenAI, DefaultHttpxClient
855+
import httpx2
856+
from openai import OpenAI, DefaultHttpx2Client
885857

886858
client = OpenAI(
887859
# Or use the `OPENAI_BASE_URL` env var
888860
base_url="http://my.test.server.example.com:8083/v1",
889-
http_client=DefaultHttpxClient(
861+
http_client=DefaultHttpx2Client(
890862
proxy="http://my.test.proxy.example.com",
891-
transport=httpx.HTTPTransport(local_address="0.0.0.0"),
863+
transport=httpx2.HTTPTransport(local_address="0.0.0.0"),
892864
),
893865
)
894866
```
895867

896868
You can also customize the client on a per-request basis by using `with_options()`:
897869

898870
```python
899-
client.with_options(http_client=DefaultHttpxClient(...))
871+
client.with_options(http_client=DefaultHttpx2Client(...))
900872
```
901873

902874
#### Mutual TLS
@@ -913,7 +885,7 @@ and pass it through the custom HTTP client:
913885
import os
914886
import ssl
915887

916-
from openai import OpenAI, DefaultHttpxClient
888+
from openai import OpenAI, DefaultHttpx2Client
917889

918890
# Server trust is configured independently. Without `cafile`, this uses the
919891
# operating system's normal trusted certificate authorities.
@@ -938,7 +910,7 @@ client = OpenAI(
938910
),
939911
# A client certificate belongs to the HTTP client, not the base URL.
940912
# Disable redirects so it cannot follow a response to another origin.
941-
http_client=DefaultHttpxClient(
913+
http_client=DefaultHttpx2Client(
942914
verify=ssl_context,
943915
follow_redirects=False,
944916
),
@@ -951,7 +923,7 @@ The async configuration is equivalent:
951923
import os
952924
import ssl
953925

954-
from openai import AsyncOpenAI, DefaultAsyncHttpxClient
926+
from openai import AsyncOpenAI, DefaultAsyncHttpx2Client
955927

956928
ssl_context = ssl.create_default_context(
957929
cafile=os.environ.get("OPENAI_MTLS_CA_BUNDLE"),
@@ -968,27 +940,7 @@ client = AsyncOpenAI(
968940
"OPENAI_BASE_URL",
969941
"https://mtls.api.openai.com/v1",
970942
),
971-
http_client=DefaultAsyncHttpxClient(
972-
verify=ssl_context,
973-
follow_redirects=False,
974-
),
975-
)
976-
```
977-
978-
Experimental HTTPX2 uses the same native `SSLContext`. Install the optional
979-
extra with `pip install 'openai[httpx2]'`, then use `DefaultHttpx2Client` or
980-
`DefaultAsyncHttpx2Client` in place of the corresponding HTTPX client above:
981-
982-
```python
983-
from openai import OpenAI, DefaultHttpx2Client
984-
985-
client = OpenAI(
986-
api_key=os.environ["OPENAI_API_KEY"],
987-
base_url=os.environ.get(
988-
"OPENAI_BASE_URL",
989-
"https://mtls.api.openai.com/v1",
990-
),
991-
http_client=DefaultHttpx2Client(
943+
http_client=DefaultAsyncHttpx2Client(
992944
verify=ssl_context,
993945
follow_redirects=False,
994946
),
@@ -1001,7 +953,7 @@ See the complete [sync HTTPX2](examples/mtls_httpx2.py) and
1001953
The certificate-bearing HTTP client is transport-wide. Dedicate it to the
1002954
selected mTLS origin; do not reuse it for other services or pass it through
1003955
`with_options()` with a different `base_url`. If redirects are required, add an
1004-
HTTPX request hook that rejects requests whose scheme, host, or port differs
956+
HTTPX2 request hook that rejects requests whose scheme, host, or port differs
1005957
from the configured mTLS origin before enabling `follow_redirects`.
1006958

1007959
`SSLContext.load_cert_chain()` raises during setup for unreadable or malformed

0 commit comments

Comments
 (0)