Groq
Login
Loading model information...

Key Technical Specifications

Model Architecture

Compound mini is powered by Llama 3.3 70B and GPT-OSS 120B for intelligent reasoning and tool use. Unlike groq/compound, it can only use one tool per request, but has an average of 3x lower latency.

Performance Metrics

Groq developed a new evaluation benchmark for measuring search capabilities called RealtimeEval. This benchmark is designed to evaluate tool-using systems on current events and live data. On the benchmark, Compound Mini outperformed GPT-4o-search-preview and GPT-4o-mini-search-preview significantly.
Learn More About Agentic Tooling
Discover how to build powerful applications with real-time web search and code execution

Use Cases

Realtime Web Search

Automatically access up-to-date information from the web using the built-in web search tool.

Code Execution

Execute Python code automatically using the code execution tool powered by E2B.

Code Generation and Technical Tasks
Create AI tools for code generation, debugging, and technical problem-solving with high-quality multilingual support.

Best Practices

  • Use system prompts to improve steerability and reduce false refusals. Compound mini is designed to be highly steerable with appropriate system prompts.
  • Consider implementing system-level protections like Llama Guard for input filtering and response validation.
  • Deploy with appropriate safeguards when working in specialized domains or with critical content.

Quick Start

Experience the capabilities of groq/compound-mini on Groq:

shell
pip install groq
Python
from groq import Groq
client = Groq()
completion = client.chat.completions.create(
    model="groq/compound-mini",
    messages=[
        {
            "role": "user",
            "content": "Explain why fast inference is critical for reasoning models"
        }
    ]
)
print(completion.choices[0].message.content)

Was this page helpful?