feat: async llm support

feat: introduce async contract to BaseLLM

feat: add async call support for:

Azure provider

Anthropic provider

OpenAI provider

Gemini provider

Bedrock provider

LiteLLM provider

chore: expand scrubbed header fields (conftest, anthropic, bedrock)

chore: update docs to cover async functionality

chore: update and harden tests to support acall; re-add uri for cassette compatibility

chore: generate missing cassette

fix: ensure acall is non-abstract and set supports_tools = true for supported Anthropic models

chore: improve Bedrock async docstring and general test robustness

lib/crewai/tests/llms/openai/test_openai.py

@@ -475,10 +475,14 @@ def test_openai_get_client_params_priority_order():
         params3 = llm3._get_client_params()
         assert params3["base_url"] == "https://env.openai.com/v1"
 
-def test_openai_get_client_params_no_base_url():
+def test_openai_get_client_params_no_base_url(monkeypatch):
     """
     Test that _get_client_params works correctly when no base_url is specified
     """
+    # Clear env vars that could set base_url
+    monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
+    monkeypatch.delenv("OPENAI_API_BASE", raising=False)
+
     llm = OpenAICompletion(model="gpt-4o")
     client_params = llm._get_client_params()
     # When no base_url is provided, it should not be in the params (filtered out as None)

lib/crewai/tests/llms/openai/test_openai_async.py

@@ -0,0 +1,139 @@
+"""Tests for OpenAI async completion functionality."""
+
+import pytest
+import tiktoken
+
+from crewai.llm import LLM
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_basic_call():
+    """Test basic async call with OpenAI."""
+    llm = LLM(model="gpt-4o-mini")
+
+    result = await llm.acall("Say hello")
+
+    assert result is not None
+    assert isinstance(result, str)
+    assert len(result) > 0
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_with_temperature():
+    """Test async call with temperature parameter."""
+    llm = LLM(model="gpt-4o-mini", temperature=0.1)
+
+    result = await llm.acall("Say the word 'test' once")
+
+    assert result is not None
+    assert isinstance(result, str)
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_with_max_tokens():
+    """Test async call with max_tokens parameter."""
+    llm = LLM(model="gpt-4o-mini", max_tokens=10)
+
+    result = await llm.acall("Write a very long story about a dragon.")
+
+    assert result is not None
+    assert isinstance(result, str)
+    encoder = tiktoken.get_encoding("cl100k_base")
+    token_count = len(encoder.encode(result))
+    assert token_count <= 10
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_with_system_message():
+    """Test async call with system message."""
+    llm = LLM(model="gpt-4o-mini")
+
+    messages = [
+        {"role": "system", "content": "You are a helpful assistant."},
+        {"role": "user", "content": "What is 2+2?"}
+    ]
+
+    result = await llm.acall(messages)
+
+    assert result is not None
+    assert isinstance(result, str)
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_conversation():
+    """Test async call with conversation history."""
+    llm = LLM(model="gpt-4o-mini")
+
+    messages = [
+        {"role": "user", "content": "My name is Alice."},
+        {"role": "assistant", "content": "Hello Alice! Nice to meet you."},
+        {"role": "user", "content": "What is my name?"}
+    ]
+
+    result = await llm.acall(messages)
+
+    assert result is not None
+    assert isinstance(result, str)
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_multiple_calls():
+    """Test making multiple async calls in sequence."""
+    llm = LLM(model="gpt-4o-mini")
+
+    result1 = await llm.acall("What is 1+1?")
+    result2 = await llm.acall("What is 2+2?")
+
+    assert result1 is not None
+    assert result2 is not None
+    assert isinstance(result1, str)
+    assert isinstance(result2, str)
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_with_response_format_none():
+    """Test async call with response_format set to None."""
+    llm = LLM(model="gpt-4o-mini", response_format=None)
+
+    result = await llm.acall("Tell me a short fact")
+
+    assert result is not None
+    assert isinstance(result, str)
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_with_response_format_json():
+    """Test async call with JSON response format."""
+    llm = LLM(model="gpt-4o-mini", response_format={"type": "json_object"})
+
+    result = await llm.acall("Return a JSON object with a 'greeting' field")
+
+    assert result is not None
+    assert isinstance(result, str)
+
+
+@pytest.mark.vcr()
+@pytest.mark.asyncio
+async def test_openai_async_with_parameters():
+    """Test async call with multiple parameters."""
+    llm = LLM(
+        model="gpt-4o-mini",
+        temperature=0.7,
+        max_tokens=100,
+        top_p=0.9,
+        frequency_penalty=0.5,
+        presence_penalty=0.3
+    )
+
+    result = await llm.acall("Tell me a short fact")
+
+    assert result is not None
+    assert isinstance(result, str)