Tune Blob

Blob is a powerhouse AI assistant model designed for developers who need advanced, out-of-the-box capabilities. It’s not just another language model – it’s a complete solution that combines multiple LLMs, web search, function calling, image generation, and more, all accessible through a single model.

Blob combines multiple language models in the background, offering significantly improved intelligence over standard models. It’s designed to be flexible and adaptive, with its underlying models changing based on performance.

​

Key Features

• 10x Smarter: Simply changing the model ID grants access to greatly enhanced capabilities.
• Dynamic Model Selection: Not limited to currently available models on Tune – adapts based on performance.
• Comprehensive Capabilities: Includes web search, image generation, YouTube transcription, function calling, and email functionality.
• Assistant API Backend: Built on the same technology as the Assistant API, allowing users to recreate Blob’s functionality.

​

How to Use

To use Tune Blob, make a chat completion API call and set the model ID to kaushikaakash04/tune-blob in your request body.

​

Example

import requests

url = "https://proxy.tune.app/chat/completions"
headers = {
    "Authorization": "sk-tune-Hxxx40Ozu",
    "Content-Type": "application/json",
}
data = {
    "temperature": 0.8,
    "messages": [
        {
            "role": "user",
            "content": "Research the latest trends in AI and summarize them in bullet points"
        }
    ],
    "model": "kaushikaakash04/tune-blob",
    "stream": False,
    "frequency_penalty": 0,
    "max_tokens": 900
}

response = requests.post(url, headers=headers, json=data)
print(response.json())

This code sends a request to Blob, asking it to research AI trends and summarize them. Tune Blob will use its internet access and other tools to generate a response.

​

Advanced Capabilities

​

Web Search

• Searches multiple pages in parallel until the context window is filled.
• Provides maximum amount of results (up to 10 pages).

​

Email Functionality

• Available only in chat mode.
• Allows sending emails to the user’s registered email address.
• Designed to prevent system abuse by limiting recipients.

​

Function Calling

• Integrated by default, enhancing the model’s ability to perform specific tasks.

​

Performance Considerations

When using Tune Blob, consider the trade-off between:

• Slow and accurate responses
• Fast but potentially less accurate responses

​

Current Model Composition

Tune Blob currently utilizes a combination of:

• OpenAI GPT Models (internally uses gpt-4o, and gpt-3.5)
• Anthropic’s Claude (including Haiku version)
• Various open-source models

​

Limitations

• Email functionality is limited to sending emails to the user’s own registered email address.
• The model’s composition and capabilities may change without notice as we continuously improve the service.

For more detailed information on API usage and integration, please refer to our API Reference.