# Azure Switchboard
Batteries-included, coordination-free client loadbalancing for Azure OpenAI.
```bash
uv add azure-switchboard
```
[](https://pypi.org/project/azure-switchboard/)
[](https://opensource.org/licenses/MIT)
[](https://github.com/arini-ai/azure-switchboard/actions/workflows/ci.yaml)
## Overview
`azure-switchboard` is a Python 3 asyncio library that provides an intelligent, API-compatible client loadbalancer for Azure OpenAI. You instantiate a Switchboard client with a set of deployments, and the client distributes your chat completion requests across the available deployments using the [power of two random choices](https://www.eecs.harvard.edu/~michaelm/postscripts/handbook2001.pdf) method. In this sense, it functions as a lightweight service mesh between your application and Azure OpenAI. The basic idea is inspired by [ServiceRouter](https://www.usenix.org/system/files/osdi23-saokar.pdf).
## Features
- **API Compatibility**: `Switchboard.create` is a transparently-typed drop-in proxy for `OpenAI.chat.completions.create`.
- **Coordination-Free**: The default Two Random Choices algorithm does not require coordination between client instances to achieve excellent load distribution characteristics.
- **Utilization-Aware**: TPM/RPM ratelimit utilization is tracked per model per deployment for use during selection.
- **Batteries Included**:
- **Session Affinity**: Provide a `session_id` to route requests in the same session to the same deployment, optimizing for prompt caching
- **Automatic Failover**: Client automatically retries on request failure, with optional fallback to OpenAI by providing an `OpenAIDeployment` in `deployments`. The retry policy can also be customized by passing a tenacity
`AsyncRetrying` instance to `failover_policy`.
- **Pluggable Selection**: Custom selection algorithms can be
provided by passing a callable to the `selector` parameter on the Switchboard constructor.
- **OpenTelemetry Integration**: Comprehensive metrics and instrumentation for monitoring deployment health and utilization.
- **Lightweight**: sub-400 LOC implementation with minimal dependencies: `openai`, `tenacity`, `wrapt`, and `opentelemetry-api`. <1ms overhead per request.
- **100% Test Coverage**: Comprehensive test suite with pytest.
## Runnable Example
```python
#!/usr/bin/env python3
#
# To run this, use:
# uv run --env-file .env tools/readme_example.py
#
# /// script
# requires-python = ">=3.10"
# dependencies = [
# "azure-switchboard",
# ]
# ///
import asyncio
import os
from azure_switchboard import AzureDeployment, Model, OpenAIDeployment, Switchboard
azure_openai_endpoint = os.getenv("AZURE_OPENAI_ENDPOINT")
azure_openai_api_key = os.getenv("AZURE_OPENAI_API_KEY")
openai_api_key = os.getenv("OPENAI_API_KEY", None)
deployments = []
if azure_openai_endpoint and azure_openai_api_key:
# create 3 deployments. reusing the endpoint
# is fine for the purposes of this demo
for name in ("east", "west", "south"):
deployments.append(
AzureDeployment(
name=name,
endpoint=azure_openai_endpoint,
api_key=azure_openai_api_key,
models=[Model(name="gpt-4o-mini")],
)
)
if openai_api_key:
# we can use openai as a fallback deployment
# it will pick up the api key from the environment
deployments.append(OpenAIDeployment())
async def main():
async with Switchboard(deployments=deployments) as sb:
print("Basic functionality:")
await basic_functionality(sb)
print("Session affinity (should warn):")
await session_affinity(sb)
async def basic_functionality(switchboard: Switchboard):
# Make a completion request (non-streaming)
response = await switchboard.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Hello, world!"}],
)
print("completion:", response.choices[0].message.content)
# Make a streaming completion request
stream = await switchboard.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Hello, world!"}],
stream=True,
)
print("streaming: ", end="")
async for chunk in stream:
if chunk.choices and chunk.choices[0].delta.content:
print(chunk.choices[0].delta.content, end="", flush=True)
print()
async def session_affinity(switchboard: Switchboard):
session_id = "anything"
# First message will select a random healthy
# deployment and associate it with the session_id
r = await switchboard.create(
session_id=session_id,
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Who won the World Series in 2020?"}],
)
d1 = switchboard.select_deployment(model="gpt-4o-mini", session_id=session_id)
print("deployment 1:", d1)
print("response 1:", r.choices[0].message.content)
# Follow-up requests with the same session_id will route to the same deployment
r2 = await switchboard.create(
session_id=session_id,
model="gpt-4o-mini",
messages=[
{"role": "user", "content": "Who won the World Series in 2020?"},
{"role": "assistant", "content": r.choices[0].message.content},
{"role": "user", "content": "Who did they beat?"},
],
)
print("response 2:", r2.choices[0].message.content)
# Simulate a failure by marking down the deployment
d1.models["gpt-4o-mini"].cooldown()
# A new deployment will be selected for this session_id
r3 = await switchboard.create(
session_id=session_id,
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Who won the World Series in 2021?"}],
)
d2 = switchboard.select_deployment(model="gpt-4o-mini", session_id=session_id)
print("deployment 2:", d2)
print("response 3:", r3.choices[0].message.content)
assert d2 != d1
if __name__ == "__main__":
asyncio.run(main())
```
## Benchmarks
```bash
just bench
uv run --env-file .env tools/bench.py -v -r 1000 -d 10 -e 500
Distributing 1000 requests across 10 deployments
Max inflight requests: 1000
Request 500/1000 completed
Utilization Distribution:
0.000 - 0.200 | 0
0.200 - 0.400 | 10 ..............................
0.400 - 0.600 | 0
0.600 - 0.800 | 0
0.800 - 1.000 | 0
Avg utilization: 0.339 (0.332 - 0.349)
Std deviation: 0.006
{
'bench_0': {'gpt-4o-mini': {'util': 0.361, 'tpm': '10556/30000', 'rpm': '100/300'}},
'bench_1': {'gpt-4o-mini': {'util': 0.339, 'tpm': '9819/30000', 'rpm': '100/300'}},
'bench_2': {'gpt-4o-mini': {'util': 0.333, 'tpm': '9405/30000', 'rpm': '97/300'}},
'bench_3': {'gpt-4o-mini': {'util': 0.349, 'tpm': '10188/30000', 'rpm': '100/300'}},
'bench_4': {'gpt-4o-mini': {'util': 0.346, 'tpm': '10210/30000', 'rpm': '99/300'}},
'bench_5': {'gpt-4o-mini': {'util': 0.341, 'tpm': '10024/30000', 'rpm': '99/300'}},
'bench_6': {'gpt-4o-mini': {'util': 0.343, 'tpm': '10194/30000', 'rpm': '100/300'}},
'bench_7': {'gpt-4o-mini': {'util': 0.352, 'tpm': '10362/30000', 'rpm': '102/300'}},
'bench_8': {'gpt-4o-mini': {'util': 0.35, 'tpm': '10362/30000', 'rpm': '102/300'}},
'bench_9': {'gpt-4o-mini': {'util': 0.365, 'tpm': '10840/30000', 'rpm': '101/300'}}
}
Utilization Distribution:
0.000 - 0.100 | 0
0.100 - 0.200 | 0
0.200 - 0.300 | 0
0.300 - 0.400 | 10 ..............................
0.400 - 0.500 | 0
0.500 - 0.600 | 0
0.600 - 0.700 | 0
0.700 - 0.800 | 0
0.800 - 0.900 | 0
0.900 - 1.000 | 0
Avg utilization: 0.348 (0.333 - 0.365)
Std deviation: 0.009
Distribution overhead: 926.14ms
Average response latency: 5593.77ms
Total latency: 17565.37ms
Requests per second: 1079.75
Overhead per request: 0.93ms
```
Distribution overhead scales ~linearly with the number of deployments.
## Configuration Reference
### switchboard.Model Parameters
| Parameter | Description | Default |
| ------------------ | ----------------------------------------------------- | ------------- |
| `name` | Configured model name, e.g. "gpt-4o" or "gpt-4o-mini" | Required |
| `tpm` | Configured TPM rate limit | 0 (unlimited) |
| `rpm` | Configured RPM rate limit | 0 (unlimited) |
| `default_cooldown` | Default cooldown period in seconds | 10.0 |
### switchboard.AzureDeployment Parameters
| Parameter | Description | Default |
| ------------- | --------------------------------------------- | ------------ |
| `name` | Unique identifier for the deployment | Required |
| `endpoint` | Azure OpenAI endpoint URL | Required |
| `api_key` | Azure OpenAI API key | Required |
| `api_version` | Azure OpenAI API version | "2024-10-21" |
| `timeout` | Default timeout in seconds | 600.0 |
| `models` | List of Models configured for this deployment | Required |
### switchboard.Switchboard Parameters
| Parameter | Description | Default |
| ------------------ | ----------------------------------- | ------------------------------------------- |
| `deployments` | List of Deployment config objects | Required |
| `selector` | Selection algorithm | `two_random_choices` |
| `failover_policy` | Policy for handling failed requests | `AsyncRetrying(stop=stop_after_attempt(2))` |
| `ratelimit_window` | Ratelimit window in seconds | 60.0 |
| `max_sessions` | Maximum number of sessions | 1024 |
## Development
This project uses [uv](https://github.com/astral-sh/uv) for package management,
and [just](https://github.com/casey/just) for task automation. See the [justfile](https://github.com/arini-ai/azure-switchboard/blob/master/justfile)
for available commands.
```bash
git clone https://github.com/arini-ai/azure-switchboard
cd azure-switchboard
just install
```
### Running tests
```bash
just test
```
### Release
This library uses CalVer for versioning. On push to master, if tests pass, a package is automatically built, released, and uploaded to PyPI.
Locally, the package can be built with uv:
```bash
uv build
```
### OpenTelemetry Integration
The library provides instrumentation for monitoring deployment health and performance metrics:
```bash
(azure-switchboard) .venv > just otel-run
uv run --env-file .env opentelemetry-instrument python tools/bench.py -r 5 -d 3
Distributing 5 requests across 3 deployments
Max inflight requests: 1000
Distribution overhead: 10.53ms
Average response latency: 2164.03ms
Total latency: 3869.06ms
Requests per second: 475.03
Overhead per request: 2.11ms
{
"resource_metrics": [
{
"resource": {
"attributes": {
"telemetry.sdk.language": "python",
"telemetry.sdk.name": "opentelemetry",
"telemetry.sdk.version": "1.31.0",
"service.name": "switchboard",
"telemetry.auto.version": "0.52b0"
},
"schema_url": ""
},
"scope_metrics": [
{
"scope": {
"name": "azure_switchboard.deployment",
"version": "",
"schema_url": "",
"attributes": null
},
"metrics": [
{
"name": "model_utilization",
"description": "Current utilization of a model deployment (0-1)",
"unit": "percent",
"data": {
"data_points": [
{
"attributes": {
"model": "gpt-4o-mini"
},
"start_time_unix_nano": null,
"time_unix_nano": 1742461487509982000,
"value": 0.008,
"exemplars": []
...
```
## Contributing
1. Fork/clone repo
2. Make changes
3. Run tests with `just test`
4. Lint with `just lint`
5. Commit and make a PR
## License
MIT
Raw data
{
"_id": null,
"home_page": null,
"name": "azure-switchboard",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.10",
"maintainer_email": null,
"keywords": "ai, azure, litellm, llm, loadbalancing, openai",
"author": null,
"author_email": "Abizer Lokhandwala <abizer@abizer.me>",
"download_url": "https://files.pythonhosted.org/packages/3b/39/3b3ffbe0f1ec053a8a93593ad0a22bf26d6c03437c90deeacc0294d1556d/azure_switchboard-2025.8.0.tar.gz",
"platform": null,
"description": "# Azure Switchboard\n\nBatteries-included, coordination-free client loadbalancing for Azure OpenAI.\n\n```bash\nuv add azure-switchboard\n```\n\n[](https://pypi.org/project/azure-switchboard/)\n[](https://opensource.org/licenses/MIT)\n[](https://github.com/arini-ai/azure-switchboard/actions/workflows/ci.yaml)\n\n## Overview\n\n`azure-switchboard` is a Python 3 asyncio library that provides an intelligent, API-compatible client loadbalancer for Azure OpenAI. You instantiate a Switchboard client with a set of deployments, and the client distributes your chat completion requests across the available deployments using the [power of two random choices](https://www.eecs.harvard.edu/~michaelm/postscripts/handbook2001.pdf) method. In this sense, it functions as a lightweight service mesh between your application and Azure OpenAI. The basic idea is inspired by [ServiceRouter](https://www.usenix.org/system/files/osdi23-saokar.pdf).\n\n## Features\n\n- **API Compatibility**: `Switchboard.create` is a transparently-typed drop-in proxy for `OpenAI.chat.completions.create`.\n- **Coordination-Free**: The default Two Random Choices algorithm does not require coordination between client instances to achieve excellent load distribution characteristics.\n- **Utilization-Aware**: TPM/RPM ratelimit utilization is tracked per model per deployment for use during selection.\n- **Batteries Included**:\n - **Session Affinity**: Provide a `session_id` to route requests in the same session to the same deployment, optimizing for prompt caching\n - **Automatic Failover**: Client automatically retries on request failure, with optional fallback to OpenAI by providing an `OpenAIDeployment` in `deployments`. The retry policy can also be customized by passing a tenacity\n `AsyncRetrying` instance to `failover_policy`.\n - **Pluggable Selection**: Custom selection algorithms can be\n provided by passing a callable to the `selector` parameter on the Switchboard constructor.\n - **OpenTelemetry Integration**: Comprehensive metrics and instrumentation for monitoring deployment health and utilization.\n\n- **Lightweight**: sub-400 LOC implementation with minimal dependencies: `openai`, `tenacity`, `wrapt`, and `opentelemetry-api`. <1ms overhead per request.\n- **100% Test Coverage**: Comprehensive test suite with pytest.\n\n## Runnable Example\n\n```python\n#!/usr/bin/env python3\n#\n# To run this, use:\n# uv run --env-file .env tools/readme_example.py\n#\n# /// script\n# requires-python = \">=3.10\"\n# dependencies = [\n# \"azure-switchboard\",\n# ]\n# ///\n\nimport asyncio\nimport os\n\nfrom azure_switchboard import AzureDeployment, Model, OpenAIDeployment, Switchboard\n\nazure_openai_endpoint = os.getenv(\"AZURE_OPENAI_ENDPOINT\")\nazure_openai_api_key = os.getenv(\"AZURE_OPENAI_API_KEY\")\nopenai_api_key = os.getenv(\"OPENAI_API_KEY\", None)\n\ndeployments = []\nif azure_openai_endpoint and azure_openai_api_key:\n # create 3 deployments. reusing the endpoint\n # is fine for the purposes of this demo\n for name in (\"east\", \"west\", \"south\"):\n deployments.append(\n AzureDeployment(\n name=name,\n endpoint=azure_openai_endpoint,\n api_key=azure_openai_api_key,\n models=[Model(name=\"gpt-4o-mini\")],\n )\n )\n\nif openai_api_key:\n # we can use openai as a fallback deployment\n # it will pick up the api key from the environment\n deployments.append(OpenAIDeployment())\n\n\nasync def main():\n async with Switchboard(deployments=deployments) as sb:\n print(\"Basic functionality:\")\n await basic_functionality(sb)\n\n print(\"Session affinity (should warn):\")\n await session_affinity(sb)\n\n\nasync def basic_functionality(switchboard: Switchboard):\n # Make a completion request (non-streaming)\n response = await switchboard.create(\n model=\"gpt-4o-mini\",\n messages=[{\"role\": \"user\", \"content\": \"Hello, world!\"}],\n )\n\n print(\"completion:\", response.choices[0].message.content)\n\n # Make a streaming completion request\n stream = await switchboard.create(\n model=\"gpt-4o-mini\",\n messages=[{\"role\": \"user\", \"content\": \"Hello, world!\"}],\n stream=True,\n )\n\n print(\"streaming: \", end=\"\")\n async for chunk in stream:\n if chunk.choices and chunk.choices[0].delta.content:\n print(chunk.choices[0].delta.content, end=\"\", flush=True)\n\n print()\n\n\nasync def session_affinity(switchboard: Switchboard):\n session_id = \"anything\"\n\n # First message will select a random healthy\n # deployment and associate it with the session_id\n r = await switchboard.create(\n session_id=session_id,\n model=\"gpt-4o-mini\",\n messages=[{\"role\": \"user\", \"content\": \"Who won the World Series in 2020?\"}],\n )\n\n d1 = switchboard.select_deployment(model=\"gpt-4o-mini\", session_id=session_id)\n print(\"deployment 1:\", d1)\n print(\"response 1:\", r.choices[0].message.content)\n\n # Follow-up requests with the same session_id will route to the same deployment\n r2 = await switchboard.create(\n session_id=session_id,\n model=\"gpt-4o-mini\",\n messages=[\n {\"role\": \"user\", \"content\": \"Who won the World Series in 2020?\"},\n {\"role\": \"assistant\", \"content\": r.choices[0].message.content},\n {\"role\": \"user\", \"content\": \"Who did they beat?\"},\n ],\n )\n\n print(\"response 2:\", r2.choices[0].message.content)\n\n # Simulate a failure by marking down the deployment\n d1.models[\"gpt-4o-mini\"].cooldown()\n\n # A new deployment will be selected for this session_id\n r3 = await switchboard.create(\n session_id=session_id,\n model=\"gpt-4o-mini\",\n messages=[{\"role\": \"user\", \"content\": \"Who won the World Series in 2021?\"}],\n )\n\n d2 = switchboard.select_deployment(model=\"gpt-4o-mini\", session_id=session_id)\n print(\"deployment 2:\", d2)\n print(\"response 3:\", r3.choices[0].message.content)\n assert d2 != d1\n\n\nif __name__ == \"__main__\":\n asyncio.run(main())\n```\n\n## Benchmarks\n\n```bash\njust bench\nuv run --env-file .env tools/bench.py -v -r 1000 -d 10 -e 500\nDistributing 1000 requests across 10 deployments\nMax inflight requests: 1000\n\nRequest 500/1000 completed\nUtilization Distribution:\n0.000 - 0.200 | 0\n0.200 - 0.400 | 10 ..............................\n0.400 - 0.600 | 0\n0.600 - 0.800 | 0\n0.800 - 1.000 | 0\nAvg utilization: 0.339 (0.332 - 0.349)\nStd deviation: 0.006\n\n{\n 'bench_0': {'gpt-4o-mini': {'util': 0.361, 'tpm': '10556/30000', 'rpm': '100/300'}},\n 'bench_1': {'gpt-4o-mini': {'util': 0.339, 'tpm': '9819/30000', 'rpm': '100/300'}},\n 'bench_2': {'gpt-4o-mini': {'util': 0.333, 'tpm': '9405/30000', 'rpm': '97/300'}},\n 'bench_3': {'gpt-4o-mini': {'util': 0.349, 'tpm': '10188/30000', 'rpm': '100/300'}},\n 'bench_4': {'gpt-4o-mini': {'util': 0.346, 'tpm': '10210/30000', 'rpm': '99/300'}},\n 'bench_5': {'gpt-4o-mini': {'util': 0.341, 'tpm': '10024/30000', 'rpm': '99/300'}},\n 'bench_6': {'gpt-4o-mini': {'util': 0.343, 'tpm': '10194/30000', 'rpm': '100/300'}},\n 'bench_7': {'gpt-4o-mini': {'util': 0.352, 'tpm': '10362/30000', 'rpm': '102/300'}},\n 'bench_8': {'gpt-4o-mini': {'util': 0.35, 'tpm': '10362/30000', 'rpm': '102/300'}},\n 'bench_9': {'gpt-4o-mini': {'util': 0.365, 'tpm': '10840/30000', 'rpm': '101/300'}}\n}\n\nUtilization Distribution:\n0.000 - 0.100 | 0\n0.100 - 0.200 | 0\n0.200 - 0.300 | 0\n0.300 - 0.400 | 10 ..............................\n0.400 - 0.500 | 0\n0.500 - 0.600 | 0\n0.600 - 0.700 | 0\n0.700 - 0.800 | 0\n0.800 - 0.900 | 0\n0.900 - 1.000 | 0\nAvg utilization: 0.348 (0.333 - 0.365)\nStd deviation: 0.009\n\nDistribution overhead: 926.14ms\nAverage response latency: 5593.77ms\nTotal latency: 17565.37ms\nRequests per second: 1079.75\nOverhead per request: 0.93ms\n```\n\nDistribution overhead scales ~linearly with the number of deployments.\n\n## Configuration Reference\n\n### switchboard.Model Parameters\n\n| Parameter | Description | Default |\n| ------------------ | ----------------------------------------------------- | ------------- |\n| `name` | Configured model name, e.g. \"gpt-4o\" or \"gpt-4o-mini\" | Required |\n| `tpm` | Configured TPM rate limit | 0 (unlimited) |\n| `rpm` | Configured RPM rate limit | 0 (unlimited) |\n| `default_cooldown` | Default cooldown period in seconds | 10.0 |\n\n### switchboard.AzureDeployment Parameters\n\n| Parameter | Description | Default |\n| ------------- | --------------------------------------------- | ------------ |\n| `name` | Unique identifier for the deployment | Required |\n| `endpoint` | Azure OpenAI endpoint URL | Required |\n| `api_key` | Azure OpenAI API key | Required |\n| `api_version` | Azure OpenAI API version | \"2024-10-21\" |\n| `timeout` | Default timeout in seconds | 600.0 |\n| `models` | List of Models configured for this deployment | Required |\n\n### switchboard.Switchboard Parameters\n\n| Parameter | Description | Default |\n| ------------------ | ----------------------------------- | ------------------------------------------- |\n| `deployments` | List of Deployment config objects | Required |\n| `selector` | Selection algorithm | `two_random_choices` |\n| `failover_policy` | Policy for handling failed requests | `AsyncRetrying(stop=stop_after_attempt(2))` |\n| `ratelimit_window` | Ratelimit window in seconds | 60.0 |\n| `max_sessions` | Maximum number of sessions | 1024 |\n\n## Development\n\nThis project uses [uv](https://github.com/astral-sh/uv) for package management,\nand [just](https://github.com/casey/just) for task automation. See the [justfile](https://github.com/arini-ai/azure-switchboard/blob/master/justfile)\nfor available commands.\n\n```bash\ngit clone https://github.com/arini-ai/azure-switchboard\ncd azure-switchboard\n\njust install\n```\n\n### Running tests\n\n```bash\njust test\n```\n\n### Release\n\nThis library uses CalVer for versioning. On push to master, if tests pass, a package is automatically built, released, and uploaded to PyPI.\n\nLocally, the package can be built with uv:\n\n```bash\nuv build\n```\n\n### OpenTelemetry Integration\n\nThe library provides instrumentation for monitoring deployment health and performance metrics:\n\n```bash\n(azure-switchboard) .venv > just otel-run\nuv run --env-file .env opentelemetry-instrument python tools/bench.py -r 5 -d 3\nDistributing 5 requests across 3 deployments\nMax inflight requests: 1000\n\nDistribution overhead: 10.53ms\nAverage response latency: 2164.03ms\nTotal latency: 3869.06ms\nRequests per second: 475.03\nOverhead per request: 2.11ms\n{\n \"resource_metrics\": [\n {\n \"resource\": {\n \"attributes\": {\n \"telemetry.sdk.language\": \"python\",\n \"telemetry.sdk.name\": \"opentelemetry\",\n \"telemetry.sdk.version\": \"1.31.0\",\n \"service.name\": \"switchboard\",\n \"telemetry.auto.version\": \"0.52b0\"\n },\n \"schema_url\": \"\"\n },\n \"scope_metrics\": [\n {\n \"scope\": {\n \"name\": \"azure_switchboard.deployment\",\n \"version\": \"\",\n \"schema_url\": \"\",\n \"attributes\": null\n },\n \"metrics\": [\n {\n \"name\": \"model_utilization\",\n \"description\": \"Current utilization of a model deployment (0-1)\",\n \"unit\": \"percent\",\n \"data\": {\n \"data_points\": [\n {\n \"attributes\": {\n \"model\": \"gpt-4o-mini\"\n },\n \"start_time_unix_nano\": null,\n \"time_unix_nano\": 1742461487509982000,\n \"value\": 0.008,\n \"exemplars\": []\n...\n```\n\n## Contributing\n\n1. Fork/clone repo\n2. Make changes\n3. Run tests with `just test`\n4. Lint with `just lint`\n5. Commit and make a PR\n\n## License\n\nMIT\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Batteries-included loadbalancing client for Azure OpenAI",
"version": "2025.8.0",
"project_urls": {
"Homepage": "https://github.com/arini-ai/azure-switchboard"
},
"split_keywords": [
"ai",
" azure",
" litellm",
" llm",
" loadbalancing",
" openai"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "b049e662eed3b728f7ed863c7769193a19e69dba6a05ae4b45449e113435f774",
"md5": "8eff9ddf8b494d1122cc4ae4f4c9a7f8",
"sha256": "487c3cebb9298de784cee4e78b9e268ac8ee963adef6deb07950da0f1f5efc7c"
},
"downloads": -1,
"filename": "azure_switchboard-2025.8.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "8eff9ddf8b494d1122cc4ae4f4c9a7f8",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10",
"size": 12837,
"upload_time": "2025-08-07T10:23:33",
"upload_time_iso_8601": "2025-08-07T10:23:33.511531Z",
"url": "https://files.pythonhosted.org/packages/b0/49/e662eed3b728f7ed863c7769193a19e69dba6a05ae4b45449e113435f774/azure_switchboard-2025.8.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "3b393b3ffbe0f1ec053a8a93593ad0a22bf26d6c03437c90deeacc0294d1556d",
"md5": "4a34c75a35b4691a3c024d59253ec5ed",
"sha256": "724c4c5c1cbff4a821778f04cd63b54ebbc62a282d55460133ea523de576b769"
},
"downloads": -1,
"filename": "azure_switchboard-2025.8.0.tar.gz",
"has_sig": false,
"md5_digest": "4a34c75a35b4691a3c024d59253ec5ed",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10",
"size": 154612,
"upload_time": "2025-08-07T10:23:34",
"upload_time_iso_8601": "2025-08-07T10:23:34.655466Z",
"url": "https://files.pythonhosted.org/packages/3b/39/3b3ffbe0f1ec053a8a93593ad0a22bf26d6c03437c90deeacc0294d1556d/azure_switchboard-2025.8.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-07 10:23:34",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "arini-ai",
"github_project": "azure-switchboard",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "azure-switchboard"
}
JSON
