Key Technical Specifications

Best Practices

• Mode Selection: use thinking mode (reasoning_effort="default") for complex reasoning, math, and coding, and instruct mode (reasoning_effort="none") for efficient, general-purpose dialogue. Tune thinking depth with reasoning_effort="low", "medium", or "high".
• Thinking Mode: temperature=1.0, top_p=0.95, top_k=20, min_p=0.
• Instruct Mode: temperature=0.7, top_p=0.80, top_k=20, min_p=0, presence_penalty=1.5.
• Math Problems: include 'Please reason step by step, and put your final answer within \boxed{}' in the prompt.
• History Management: in multi-turn conversations, only include final outputs without thinking content.
• Reasoning Format: set reasoning_format to hidden to return only the final answer, or parsed to include the reasoning in a separate field.
• Image Inputs: each image counts as 2048 input tokens.
• Use the full 131K context window for repository-scale code and multi-document workflows.

Get Started with Qwen 3.8 27B

Experience state-of-the-art reasoning and agentic coding with Qwen 3.8 27B at Groq speed:

pip install groq

from groq import Groq
client = Groq()
completion = client.chat.completions.create(
    model="qwen/qwen3.8-27b",
    messages=[
        {
            "role": "user",
            "content": "Explain why fast inference is critical for reasoning models"
        }
    ]
)
print(completion.choices[0].message.content)