llama-index-llms-digitalocean-gradientai 0.1.5

LlamaIndex integration for DigitalOcean Gradient AI

llama-index-llms-digitalocean-gradientai

LlamaIndex integration for DigitalOcean Gradient AI with full support for function/tool calling.

Installation

pip install llama-index-llms-digitalocean-gradientai

This package uses the official gradient SDK (PyPI package: gradient) under the hood; it is installed automatically as a dependency.

Usage

Basic Usage

from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM

llm = DigitalOceanGradientAILLM(
    model="openai-gpt-oss-120b",
    model_access_key="your-api-key",
    workspace_id="your-workspace-id"  # Optional
)

response = llm.complete("What is DigitalOcean Gradient?")
print(response.text)

Chat Interface

from llama_index.core.llms import ChatMessage
from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM

llm = DigitalOceanGradientAILLM(
    model="openai-gpt-oss-120b",
    model_access_key="your-api-key",
)

messages = [
    ChatMessage(role="system", content="You are a helpful assistant."),
    ChatMessage(role="user", content="What is the capital of France?")
]

response = llm.chat(messages)
print(response.message.content)

Streaming

from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM

llm = DigitalOceanGradientAILLM(
    model="meta-llama-3-70b-instruct",
    model_access_key="your-api-key",
)

# Streaming completion
for chunk in llm.stream_complete("Tell me a story about AI:"):
    print(chunk.delta, end="", flush=True)

Async Usage

import asyncio
from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM

async def main():
    llm = DigitalOceanGradientAILLM(
        model="meta-llama-3-70b-instruct",
        model_access_key="your-api-key",
    )
    response = await llm.acomplete("What is Gradient?")
    print(response.text)

asyncio.run(main())

Function/Tool Calling

This integration supports OpenAI-compatible function calling, enabling the LLM to invoke tools based on user queries.

Using chat_with_tools

from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM
from llama_index.core.tools import FunctionTool

# Define tools
def add(a: int, b: int) -> int:
    """Add two numbers together."""
    return a + b

def multiply(a: int, b: int) -> int:
    """Multiply two numbers together."""
    return a * b

# Create tool instances
add_tool = FunctionTool.from_defaults(fn=add)
multiply_tool = FunctionTool.from_defaults(fn=multiply)
tools = [add_tool, multiply_tool]

# Initialize LLM
llm = DigitalOceanGradientAILLM(
    model="openai-gpt-oss-120b",
    model_access_key="your-api-key",
)

# Chat with tools
response = llm.chat_with_tools(
    tools=tools,
    user_msg="What is 5 multiplied by 8?",
)
print(response.message)

# Extract tool calls from response
tool_calls = llm.get_tool_calls_from_response(
    response, 
    error_on_no_tool_call=False
)
for tool_call in tool_calls:
    print(f"Tool: {tool_call.tool_name}, Args: {tool_call.tool_kwargs}")

Using predict_and_call

For automatic tool execution and result handling:

from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM
from llama_index.core.tools import FunctionTool

def add(a: int, b: int) -> int:
    """Add two numbers together."""
    return a + b

add_tool = FunctionTool.from_defaults(fn=add)

llm = DigitalOceanGradientAILLM(
    model="openai-gpt-oss-120b",
    model_access_key="your-api-key",
)

# Automatically calls the tool and returns the result
response = llm.predict_and_call(
    tools=[add_tool],
    user_msg="What is 10 plus 15?",
)
print(response)  # Output: 25

Async Function Calling

import asyncio
from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM
from llama_index.core.tools import FunctionTool

def multiply(a: int, b: int) -> int:
    """Multiply two numbers together."""
    return a * b

multiply_tool = FunctionTool.from_defaults(fn=multiply)

async def main():
    llm = DigitalOceanGradientAILLM(
        model="openai-gpt-oss-120b",
        model_access_key="your-api-key",
    )
    
    response = await llm.achat_with_tools(
        tools=[multiply_tool],
        user_msg="What is 7 times 9?",
    )
    print(response.message)

asyncio.run(main())

With RAG Pipeline

from llama_index.core import VectorStoreIndex, Document
from llama_index.llms.digitalocean.gradientai import DigitalOceanGradientAILLM

llm = DigitalOceanGradientAILLM(
    model="meta-llama-3-70b-instruct",
    model_access_key="your-api-key",
)

documents = [Document(text="DigitalOcean Gradient is a managed LLM API service...")]
index = VectorStoreIndex.from_documents(documents)
query_engine = index.as_query_engine(llm=llm)
response = query_engine.query("What is Gradient?")
print(response)

Package Structure

llama-index-llms-digitalocean-gradientai/
├── llama_index/
│   └── llms/
│       └── digitalocean/
│           └── gradientai/
│               ├── __init__.py
│               └── base.py
├── tests/
│   └── test_gradient_llm.py
├── pyproject.toml
├── README.md
└── LICENSE

License

MIT