Authenticate request using any available method.

Check if Docker is installed and running, provide installation instructions if not.

Ensure the model is available, pull it if necessary.

Execute code with MCP tools available

Synchronous PDF extraction to run in thread pool.

Extract text content from a .txt file

Extract text content from a .txt file

Get API key from environment variables.

Get API key from environment variables.

Internal method that polls the appropriate power tool until stopped.

Process an image file and return base64 encoded data or None if processing fails

ColBERT late-interaction retrieval using Liquid AI's LFM2-ColBERT-350M.

Dense embedding retrieval using SentenceTransformers + FAISS.

EmbeddingGemma retrieval using Google's EmbeddingGemma-300m model.

Gemini embedding retrieval using Google's Gemini embeddings.

MLX embedding retrieval (Apple Silicon optimized).

Multimodal retrieval using ChromaDB + Ollama.

Ollama embedding retrieval using local Ollama models.

OpenRouter embedding retrieval using OpenRouter API.

Qdrant retrieval using Qdrant + Ollama.

Run the asyncio event loop in a separate thread

Run the async chat in a new event loop in a separate thread

Run the async chat in a new event loop in a separate thread

Run the async chat in a new event loop in a separate thread

Run a single minions experiment with given parameters Args: task: Task for minions to perform contex

Handle async chat completions. Args: messages: List of message dictionaries with 'role' and 'content' keys

Asynchronously handle chat completions using the HuggingFace Inference API or Router. Args: messages: List of message di

Asynchronously handle chat completions using the HuggingFace Inference API. Args: messages: List of message dictionaries

Add two numbers and return the result.

Add entries to the ChromaDB collection. :param embeddings: list of document embeddings to add, must be either TextEmbedding, ImageEm

Add a file ID to be attached to chat messages. Example: file_info = client.upload_file("report.pdf") client.

Add a File Search store to be used in subsequent generations. Args: store_name: Name of the store to add (format

Use Firecrawl Agent to search, navigate, and gather data from the web. See: https://docs.firecrawl.dev/features/agent Example:

Get agent card.

Asynchronously handle multimodal chat completions using the Qwen2.5-Omni model. Args: messages: List of message dictiona

Asynchronously handle multimodal chat completions using the Qwen2.5-Omni model. Args: messages: List of message dictiona

Lazy-load Arch Router to avoid loading model if not needed.

Return authentication dependency function.

Get the list of available tools

Calculate required context size for given text.

Run a single timed generation to compute a correction factor and cache it. `model_client` must expose `.generate(prompt, max_tokens=.

Process messages using the Agno Agent. Args: messages: List of message dicts with 'role' and 'content' keys

Handle chat completions using the NousResearch API. Args: messages: List of message dictionaries with 'role' and 'conten

Handle chat completions via the MiniMax API. Args: messages: List of message dictionaries with 'role' and 'content' keys.

Handle chat completions via the Vercel AI Gateway. Supports sending `providerOptions` using the OpenAI SDK `extra_body` parameter. Yo

Handle chat completions using the Cloudflare AI Gateway. Args: messages: List of message dictionaries with 'role' and 'c

Handle chat completions using the OpenAI API. Args: messages: List of message dictionaries with 'role' and 'content' key

Handle chat completions using the Mistral API. If websearch is enabled, uses the websearch agent for conversations. Args:

Handle chat completions using the Modular MAX OpenAI-compatible endpoint. Args: messages: List of message dictionaries w

Handle chat completions using the Azure OpenAI API. Supports both standard models and reasoning models (o1, o3, o4).

Primary chat interface that all clients must implement. Args: messages: List of message dictionaries with 'role'

Handle chat completions with support for reasoning models. Args: messages: List of message dictionaries with 'role' and '

Handle chat completions with LMCache KV cache management. Args: messages: List of message dictionaries with 'role' and '

Main chat method: dispatches to schat or achat depending on use_async.

Handle chat completions using the Cerebras API. Args: messages: List of message dictionaries with 'role' and 'content' k

Handle chat completions using the Tencent Hunyuan API. Args: messages: List of message dictionaries with 'role' and 'con

Handle chat completions using the Osaurus server via OpenAI SDK. Args: messages: List of message dictionaries wi

Handle chat completions using the MLX Omni Server. Args: messages: List of message dictionaries with 'role' and 'content

Handle chat completions using the OpenAI API. Args: messages: List of message dictionaries with 'role' and 'content' key

Handle chat completions, routing to async or sync implementation.

Handle chat completions using the Azure OpenAI API. Args: messages: List of message dictionaries with 'role' and 'conten

Handle chat completions using the OpenAI client, but route to perplexity Args: messages: List of message dictionaries w

Handle chat completions using the Cartesia MLX client. Args: messages: List of message dictionaries with 'role' and 'con

Generate a response for a chat conversation using the MLX PARALLM model. Args: messages: List of message dictionaries, e

Handle chat completions using the OpenAI client, but route to perplexity Args: messages: List of message dictionaries w

Handle chat completions using the OpenAI API. Args: messages: List of message dictionaries with 'role' and 'content' key

Handle chat completions using the MLX LM client. Args: messages: List of message dictionaries with 'role' and 'content'

Handle chat completions using the Together API. Args: messages: List of message dictionaries with 'role' and 'content' k

Handle chat completions using the DeepSeek API. Args: messages: List of message dictionaries with 'role' and 'content' k

Handle chat completions using the Anthropic API. Args: messages: List of message dictionaries with 'role' and 'content'

Handle chat completions using the SambaNova API. Args: messages: List of message dictionaries with 'role' and 'content'

Handle chat completions using the HuggingFace Inference API. Args: messages: List of message dictionaries with 'role' an

Convenience method to perform a websearch-enabled chat with a single query. Args: query: The search query or question

Splits Python code into chunks by function (with decorators). Optionally specify number of functions per chunk

Splits Python code into chunks by function and class definitions. For stand-alone functions, the chunk is the entire function (with decorato

Clean up temporary files and old backups.

Clear all File Search stores from being used in generations. Example: client.clear_file_search_stores()

Clear all file attachments from chat (does not delete the files).

Function decorator to clear ChromaDB system cache.

Close the client.

Close the client.

Estimate the energy consumption of a GPU inference task based on approximations from epoch ai https://epoch.ai/gradient-updates/how-much-ener