> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tokenfactory.nebius.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Text-to-text

[LlamaIndex supports](https://docs.llamaindex.ai/en/latest/examples/llm/nebius/) several features for [text-to-text models](https://docs.tokenfactory.nebius.com/ai-models-inference/inference#available-models). For instance, you can send a request with a chat history or get a streaming response.

## **Prerequisites**

1. [Create an API key](https://tokenfactory.nebius.com/project/api-keys) for authentication.
2. Save the API key to an environment variable:

   ```
   export NEBIUS_API_KEY=<API_key>
   ```
3. Install LlamaIndex packages:

   ```
   pip3 install llama-index-llms-nebius llama-index
   ```

   **If you get error: externally\_managed\_ environment**

   Create a [virtual Python environment](https://docs.python.org/3/library/venv.html). You can install packages there that are isolated from the basic environment.

   To prepare a virtual environment:

   1. Create it:

      ```
      python3 -m venv <environment_name>
      ```
   2. Activate the environment:

      ```
      source <environment_name>/bin/activate
      ```

      A directory with the environment name is created.

   Now, you can install required Python packages. When you no longer need the created virtual environment, run the `deactivate` command and delete the environment directory.

## **Prepare a script**

1. Copy the following part of the script:

   ```python theme={null}
   from llama_index.llms.nebius import NebiusLLM
   import os

   # Take the API key from the environment variable
   NEBIUS_API_KEY = os.getenv("NEBIUS_API_KEY")

   # Load an LLM
   llm = NebiusLLM(
       api_key=NEBIUS_API_KEY, model="meta-llama/Meta-Llama-3.1-70B-Instruct-fast"
   )
   ```
2. Add one of the following methods, depending on your use case:

   | **Use case** | **Description** | **How to implement** |
   | :- | - | - |
   | Prompt | Ask a question as a prompt. | Call the `llm.complete("<prompt>")` method. |
   | Multi-message request | Include system prompts and a chat history to your request, so Nebius Token Factory returns more precise output. | Call the `llm.chat(messages)` method and pass along a list of messages in it.  To prepare `messages`, call the `ChatMessage()` method. For more details, see [Examples](https://docs.tokenfactory.nebius.com/integrations/frameworks/llamaIndex/text-to-text#examples). |
   | Streaming output | Output is printed out word by word. This can be helpful for chats, so the user can watch how the answer is being typed gradually. | To make the output streaming, call the `llm.stream_complete("<prompt>")` method. Next, print out the response:     **for** r **in** response:             print(r.delta, end="") |
   | Multi-message request with streaming output | Add system prompts and a chat history and receive the streaming output. | Call the `llm.stream_chat(messages)` method and pass along a list of messages in it. |

## **Examples**

### **Multi-message request**

The script below adds a chat to the request and asks in the last message to write a poem:

```python theme={null}
from llama_index.llms.nebius import NebiusLLM
from llama_index.core.llms import ChatMessage

import os

# Take the API key from the environment variable
NEBIUS_API_KEY = os.getenv("NEBIUS_API_KEY")

# Load an LLM
llm = NebiusLLM(
    api_key=NEBIUS_API_KEY, model="meta-llama/Meta-Llama-3.1-70B-Instruct-fast"
)

# Prepare a list of messages, put a request into one of them and then get a response
messages = [
    ChatMessage(role="system", content="You are a helpful AI assistant."),
    ChatMessage(
        role="user",
        content="Write a poem about a smart AI robot named Wall-e.",
    ),
]
response = llm.chat(messages)
print(response)
```

The output is the following:

```
assistant: In a world of wires and circuitry bright,
A small robot shone with a heart so light,
Wall-e, the wanderer, with a soul so true,
A smart AI mind, with a spirit anew.

...
```

### **Prompt request with streaming output**

The script below sends the prompt `Amsterdam is the capital of` and returns the output that is printed out word by word:

```python theme={null}
from llama_index.llms.nebius import NebiusLLM
from llama_index.core.llms import ChatMessage

import os

# Take the API key from the environment variable
NEBIUS_API_KEY = os.getenv("NEBIUS_API_KEY")

# Load an LLM
llm = NebiusLLM(
    api_key=NEBIUS_API_KEY, model="meta-llama/Meta-Llama-3.1-70B-Instruct-fast"
)

# Send a request as a prompt and get a response
response = llm.stream_complete("Amsterdam is the capital of ")
for r in response:
    print(r.delta, end="")
```

The output is the following:

```
The Netherlands!
```