How to adjust the model’s thinking level with reasoning_effort

The reasoning_effort parameter controls how much effort the model puts into reasoning and planning its response.

Higher effort → more thoughtful, step-by-step, or structured answers
Lower effort → faster, simpler replies

Allowed values:

Value	Behavior
`"low"`	Quick, simple replies
`"medium"`	Balanced depth and speed (good default choice)
`"high"`	More detailed reasoning and structured answers

💡 Tips:

Use "low" when you want short, direct responses.
Use "high" for tasks like explanations, problem solving, reasoning.
"medium" works well for most everyday interactions.

Note:

Setting higher reasoning effort may increase response time slightly, since the model is thinking more.

1 # Install SarvamAI
2 !pip install -Uqq sarvamai
3 from sarvamai import SarvamAI

1 # Initialize the SarvamAI client with your API key
2 client = SarvamAI(api_subscription_key="YOUR_SARVAM_API_KEY")

1 # Example 1: Using default reasoning_effort (not specified) — balanced response
2 response = client.chat.completions(
3     messages=[
4         {"role": "system", "content": "You are a helpful assistant."},
5         {"role": "user", "content": "Summarize the story of the Ramayana."}
6     ]
7     # reasoning_effort not specified → defaults internally (balanced)
8 )

1 # Example 2: Using reasoning_effort = "high" — more detailed, thoughtful response
2 response = client.chat.completions(
3     messages=[
4         {"role": "system", "content": "You are a helpful assistant."},
5         {"role": "user", "content": "Summarize the story of the Ramayana."}
6     ],
7     reasoning_effort="high"
8 )

1 # Receive assistant's reply as output.
2 print(response.choices[0].message.content)