A lightweight model that thinks before responding. Fast, smart, and great for logic-based tasks that do not require deep domain knowledge. The raw thinking traces are accessible.
At a glance
Modalities
Context window
131,072
Pricing
Capabilities
Function calling
Connect the xAI model to external tools and systems.
Structured outputs
Return responses in specific, organized formats.
Reasoning
The model can think before responding.
Pricing
Input
Tokens
$0.30/ 1M tokens
Cached tokens
$0.075/ 1M tokens
Output
Tokens
$0.50/ 1M tokens
You are charged for each token used when making calls to our API.
Using cached input tokens can significantly reduce your costs.
This model is available on multiple clusters, you can find full regional based pricing below.
Details
Model name
Aliases
| Region | us-east-1, eu-west-1 |
|---|---|
| Pricing per million tokens * | |
| Input | $0.30 |
| Cached input | $0.075 |
| Output | $0.50 |
| Rate limits | |
| Requests per minute | 1,400 |
| Tokens per minute | 4,000,000 |
Quickstart
from xai_sdk import Client
from xai_sdk.chat import user, system
client = Client(api_key="<YOUR_XAI_API_KEY_HERE>")
chat = client.chat.create(model="grok-3-mini", temperature=0)
chat.append(system("You are a PhD-level mathematician."))
chat.append(user("What is 2 + 2?"))
response = chat.sample()
print(response.content)