Retrieve entries to the ChromaDB collection based on query. :param query: embedding object (TextEmbedding) of query :param t

Retrieve information about a specific model from the Vercel AI Gateway. Args: model_id: The model identifier (e.g., "anthropic/cl

Run the minion protocol on a given query and context. Expected JSON body: { "query": "What is the main topic of this documen

Run the full minions protocol on a given task and context. Expected JSON body: { "task": "What is the main topic of this doc

Run the task using only the remote model (no minions protocol). This demonstrates the cost difference compared to the minions approach.

Perform a search using Perplexity's search capabilities. Args: query: A single query string or list of query str

Class method to perform web search using Ollama's web search API. Args: query: The search query to execute

Send a task to the A2A server.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test server with auth.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures.

Set up test fixtures before each test method.

Set up client for each test

Set up client for each test

Set up HuggingFace client

Set up Cartesia MLX client

Set up Transformers client

Set up MLX LM client

Set up LlamaCpp client

Set up MLX Omni client

Set up MLX Audio client

Set up LMCache client for each test.

Set up Secure client

Set up MLX Parallm client

Set up Lemonade client if server URL provided.

Set up Ollama client

Check API key availability before running tests

Override base class setup since distributed inference doesn't require API keys

Check LMCache dependencies before running tests.

Override to skip API key checking since Lemonade is a local service.

Set generation parameters that persist across requests.

Set the reasoning budget (maximum thinking tokens) for Nemotron models. Args: budget: Maximum tokens allowed for

Enable or disable reasoning for Nemotron models. Args: enabled: Whether to enable reasoning

Set whether to use the Interactions API for chat calls. When enabled, calls to chat() will be routed to interactions_chat().

Convert speech to text using the MLX Omni Server. Args: audio_file: File-like object containing audio data m

Convert speech to text using the MLX Omni Server. Args: audio_file: File-like object containing audio data m

Initialize components on startup.

Callback function to display story generation progress

Handle streaming chat completions using the Osaurus server. Args: messages: List of message dictionaries with 'r

Callback to send streaming updates.

Subtract b from a and return the result.

Get task status.

Clean up test fixtures.

Clean up test fixtures.

Clean up test fixtures.

Clean up test fixtures.

Clean up test fixtures.

Clean up test fixtures.

Clean up test fixtures.

Clean up LMCache resources after each test.

Test agent card retrieval.

Test agent card endpoint.

Test agent card endpoints.

Test agent card endpoints.

Test message with agent role.

Test API key security scheme.

Test basic routing decision with ArchRouter. WARNING: This test performs actual model inference and can take 10+ minutes on CPU. Set

Test batch routing of multiple queries. WARNING: This test performs multiple model inferences and can take 20+ minutes on CPU. Set RU

Test that ArchRouter correctly routes code-related tasks. WARNING: This test performs actual model inference and can take 10+ minutes on CPU.

Test that routing with empty clients raises ValueError.

Test that routing with empty query raises ValueError.

Test that ArchRouter initializes correctly. WARNING: This test loads a 1.5B parameter model and can take 2-3 minutes on CPU. Set RUN_

Test that ArchRouter generates correct XML-tagged prompt format.

Test MLX LM async mode

Test attestation verification functionality

Test authentication fails when no credentials provided.

Test optional authentication (require_auth=False).

Test authentication with API key.

Test authentication with JWT bearer token.

Test request with valid authentication.

Test authentication setup and get credentials.

Test Azure-specific configuration

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic chat functionality

Test basic Cartesia MLX chat

Test basic chat functionality

Test basic Transformers chat

Test basic MLX LM chat