The error is occurring on the Groq configuration, but the Grok API key is working fine.

[baberibrar]

Traceback (most recent call last):
  File "D:\ai-research-suite\aisuite_ai.py", line 23, in <module>
    response = client.chat.completions.create(
        model=model,
        messages=messages,
        temperature=0.75
    )
  File "D:\ai-research-suite\venv\Lib\site-packages\aisuite\client.py", line 117, in create
    return provider.chat_completions_create(model_name, messages, **kwargs)
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\ai-research-suite\venv\Lib\site-packages\aisuite\providers\groq_provider.py", line 22, in chat_completions_create
    return self.client.chat.completions.create(
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
        model=model,
        ^^^^^^^^^^^^
        messages=messages,
        ^^^^^^^^^^^^^^^^^^
        **kwargs  # Pass any additional arguments to the Groq API
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "D:\ai-research-suite\venv\Lib\site-packages\groq\resources\chat\completions.py", line 289, in create
    return self._post(
           ~~~~~~~~~~^
        "/openai/v1/chat/completions",
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<31 lines>...
        stream_cls=Stream[ChatCompletionChunk],
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "D:\ai-research-suite\venv\Lib\site-packages\groq\_base_client.py", line 1225, in post
    return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
                           ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\ai-research-suite\venv\Lib\site-packages\groq\_base_client.py", line 920, in request
    return self._request(
           ~~~~~~~~~~~~~^
        cast_to=cast_to,
        ^^^^^^^^^^^^^^^^
    ...<3 lines>...
        remaining_retries=remaining_retries,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "D:\ai-research-suite\venv\Lib\site-packages\groq\_base_client.py", line 1018, in _request
    raise self._make_status_error_from_response(err.response) from None
groq.AuthenticationError: Error code: 401 - {'error': {'message': 'Invalid API Key', 'type': 'invalid_request_error', 'code': 'invalid_api_key'}}

[rohitprasad15]

Please paste the full code on how you are loading the keys and initializing aisuite client.

[baberibrar]

Please paste the full code on how you are loading the keys and initializing aisuite client.

from dotenv import load_dotenv
import os

# Load environment variables from .env file
load_dotenv()

# Access your API keys
openai_api_key = os.getenv("OPENAI_API_KEY")
# anthropic_api_key = os.getenv("ANTHROPIC_API_KEY")
groq_api_key = os.getenv("GROQ_API_KEY")

import aisuite as ai
client = ai.Client()

models = ["groq:grok-beta"]

messages = [
    {"role": "system", "content": "Respond in Pirate English."},
    {"role": "user", "content": "Tell me a joke."},
]

for model in models:
    response = client.chat.completions.create(
        model=model,
        messages=messages,
        temperature=0.75
    )
    print(response.choices[0].message.content)

[cleesmith]

Only 2 providers work for me: OpenAI and Ollama.

I see this for Anthropic and Groq:

python -B cls1.py

Traceback (most recent call last):
  File "/Users/cleesmith/aisuite/cls1.py", line 16, in <module>
    response = client.chat.completions.create(model="groq:llama-3.1-8b-instant", messages=messages)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/cleesmith/aisuite/aisuite/client.py", line 108, in create
    self.client.providers[provider_key] = ProviderFactory.create_provider(
                                          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/cleesmith/aisuite/aisuite/provider.py", line 46, in create_provider
    return provider_class(**config)
           ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/cleesmith/aisuite/aisuite/providers/groq_provider.py", line 19, in __init__
    self.client = groq.Groq(**config)
                  ^^^^^^^^^^^^^^^^^^^
  File "/opt/miniconda3/envs/aisuite/lib/python3.11/site-packages/groq/_client.py", line 99, in __init__
    super().__init__(
  File "/opt/miniconda3/envs/aisuite/lib/python3.11/site-packages/groq/_base_client.py", line 824, in __init__
    self._client = http_client or SyncHttpxClientWrapper(
                                  ^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/miniconda3/envs/aisuite/lib/python3.11/site-packages/groq/_base_client.py", line 722, in __init__
    super().__init__(**kwargs)
TypeError: Client.__init__() got an unexpected keyword argument 'proxies'

for the following code, or the example code in this repo ... both have the same error:

# pip install 'aisuite[all]'
import aisuite as ai

client = ai.Client()

messages = [
    {"role": "system", "content": "You are a helpful agent, who answers with brevity."},
    {"role": "user", "content": 'list planets'},
]

# response = client.chat.completions.create(model="anthropic:claude-3-haiku-20240307", messages=messages)
response = client.chat.completions.create(model="groq:llama-3.1-8b-instant", messages=messages)

print(response.choices[0].message.content)

It works for openai and ollama ... using Groq is fast, free, and good for testing code so it would be nice to have.

pip show aisuite
Name: aisuite
Version: 0.1.6
Summary: Uniform access layer for LLMs
Home-page: 
Author: Andrew Ng
Author-email: 
License: 
Location: /opt/miniconda3/envs/aisuite/lib/python3.11/site-packages
Requires: 
Required-by:

[Quantisan]

related #110

[ksolo]

@baberibrar Based on the initial error it doesn't look like the env is set properly. Do you have a .env file with the keys set? You would need the GROQ_API_KEY created in that file.

I also do not see the grok-beta as a model listed on their website (https://console.groq.com/docs/models), when I attempted to make a call using that model I received a 404.

[ksolo]

@cleesmith What version of the groq client are you using?

[anandhuh1234]

Groq is working fine for me. The following models are available for inference.

import os
import aisuite as ai
aiclient = ai.Client()

os.environ['GROQ_API_KEY'] = "gsk_..."

user_input = "1+1= "

groq_models = [
    "gemma2-9b-it",
    "gemma-7b-it",
    "llama-3.1-70b-versatile",
    "llama-3.1-8b-instant",
    "llama-3.2-1b-preview",
    "llama-3.2-3b-preview",
    "llama-3.2-11b-vision-preview",
    "llama-3.2-90b-vision-preview",
    "llama-guard-3-8b",
    "llama3-70b-8192",
    "llama3-8b-8192",
    "mixtral-8x7b-32768",
]
for model in models:
	print("model: ",model)
	try:
	    response = aiclient.chat.completions.create(
				model=f"groq:{model}", 
				messages=[
					{"role": "user", "content": user_input}],
				stream=True
			)

	    # response.choices[0].message.content
	    for chunk in response:
	        data = chunk.choices[0].delta.content
	            if data is not None:
		        print(data, end='')
	except Exception as e:
		print(str(e))
	print("\n")

[ksolo]

@cleesmith I suspect your issue is the same as #110 which is caused by using httpx >= 0.28.0, in this version the proxies keyword arg was removed. If you pin your version to ~0.27.0 I would expect this to work for you. There is already a PR pinning the version to ~0.27.0. We'll need to do a round of updates on the dependencies for the python clients along with some testing before we can move to >= 0.28.0.