SwitchAI is a lightweight and flexible library that provides a standardized interface for interacting with various AI APIs like OpenAI, Anthropic, Mistral, and more. With SwitchAI, you can easily switch between AI providers or use multiple APIs simultaneously, all with a simple and consistent interface.
You can install just the base switchai
package, or install a provider's package along with it.
-
Base Package: This installs just the base
switchai
package without installing any provider's SDK.pip install switchai
-
OpenAI Provider: This installs
switchai
along with OpenAI's library.pip install switchai[openai]
-
All Providers: This installs
switchai
along with all provider-specific libraries.pip install switchai[all]
To use SwitchAI, you will need API keys for the AI providers you intend to interact with. You can set these keys either
as environment variables or pass them as configuration to the SwitchAI
client.
from switchai import SwitchAI
client = SwitchAI(provider="openai", model_name="gpt-4", api_key="your_api_key")
Set the API key as an environment variable:
macOS/Linux:
export PROVIDER_API_KEY="your_api_key"
Windows:
set PROVIDER_API_KEY="your_api_key"
Make sure you follow the correct naming conventions for each provider's API key, as outlined in the documentation. This ensures that SwitchAI can automatically detect and use the appropriate key for the chosen provider.
Here are some examples of how you can use SwitchAI to interact with different AI models:
from switchai import SwitchAI
# Initialize the client with the desired AI model
client = SwitchAI(provider="openai", model_name="gpt-4o")
# Send a message and receive a response
response = client.chat(
messages=[
{"role": "user", "content": "Hello, how are you?"}
]
)
# Print the response
print(response)
from switchai import SwitchAI
# Initialize the client with the vision model
client = SwitchAI(provider="mistral", model_name="pixtral-large-latest")
# Send an image with a question and receive a response
response = client.chat(
messages=[
{
"role": "user",
"content": [
{"type": "text", "text": "What's in this image?"},
{"type": "image", "image": "path/to/image/file.jpg"},
],
}
]
)
# Print the response
print(response)
from switchai import SwitchAI
# Initialize the client with the chosen embedding model
client = SwitchAI(provider="google", model_name="models/text-embedding-004")
# Generate embeddings for a list of text inputs
response = client.embed(
input=[
"I am feeling great today!",
"I am feeling sad today."
]
)
# Print the response
print(response)
from switchai import SwitchAI
# Initialize the client with the desired speech-to-text model
client = SwitchAI(provider="deepgram", model_name="nova-2")
# Transcribe an audio file
response = client.transcribe(
audio_path="path/to/audio/file.wav"
)
# Print the response
print(response)
from switchai import SwitchAI
client = SwitchAI(provider="replicate", model_name="black-forest-labs/flux-schnell")
response = client.generate_image("A beautiful sunset over the mountains.")
image = response.images[0]
image.show()
SuperClients are high-level interfaces that extend the base SwitchAI
client to provide additional functionalities.
Gives a chat model the ability to access websites.
from switchai import SwitchAI, Browser
client = SwitchAI(provider="openai", model_name="gpt-4o")
client = Browser(client)
response = client.chat(
messages=[
{
"role": "user",
"content": "Can you summarize the content of this website: https://example.com?"
},
]
)
print(response)
For full documentation, visit SwitchAI Documentation.
Contributions are always welcome! If you'd like to help enhance SwitchAI, feel free to make a contribution.