xAI

https://docs.x.ai/docs

tip

We support ALL xAI models, just set model=xai/<any-model-on-xai> as a prefix when sending litellm requests

Supported Models

Grok 4.5 - Frontier model for coding, agentic tasks, and knowledge work with 500K context, reasoning (low/medium/high), vision, tools, web search, and prompt caching.

Model	Context	Features
`xai/grok-4.5`	500K tokens	Reasoning, Function calling, Vision, Web search, Caching

Example:

from litellm import completion

response = completion(
    model="xai/grok-4.5",
    messages=[{"role": "user", "content": "Find and fix the bug, then explain it."}],
    reasoning_effort="high",  # low | medium | high (default high)
)

Features:

Reasoning = Chain-of-thought reasoning with reasoning tokens
Tools = Function calling / Tool use
Web search = Live internet search
Vision = Image understanding
Caching = Prompt caching for cost savings
Structured outputs = JSON / schema-constrained responses

Pricing: See xAI's pricing page for current rates.

API Key

# env variable
os.environ['XAI_API_KEY']

Sample Usage

LiteLLM python sdk usage - Non-streaming
from litellm import completion
import os

os.environ['XAI_API_KEY'] = ""
response = completion(
    model="xai/grok-4.5",
    messages=[
        {
            "role": "user",
            "content": "What's the weather like in Boston today in Fahrenheit?",
        }
    ],
    max_tokens=10,
    response_format={ "type": "json_object" },
    seed=123,
    temperature=0.2,
    top_p=0.9,
    tool_choice="auto",
    tools=[],
    user="user",
)
print(response)

Sample Usage - Streaming

LiteLLM python sdk usage - Streaming
from litellm import completion
import os

os.environ['XAI_API_KEY'] = ""
response = completion(
    model="xai/grok-4.5",
    messages=[
        {
            "role": "user",
            "content": "What's the weather like in Boston today in Fahrenheit?",
        }
    ],
    stream=True,
    max_tokens=10,
    response_format={ "type": "json_object" },
    seed=123,
    temperature=0.2,
    top_p=0.9,
    tool_choice="auto",
    tools=[],
    user="user",
)

for chunk in response:
    print(chunk)

Sample Usage - Vision

LiteLLM python sdk usage - Vision
import os 
from litellm import completion

os.environ["XAI_API_KEY"] = "your-api-key"

response = completion(
    model="xai/grok-4.5",
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://science.nasa.gov/wp-content/uploads/2023/09/web-first-images-release.png",
                        "detail": "high",
                    },
                },
                {
                    "type": "text",
                    "text": "What's in this image?",
                },
            ],
        },
    ],
)

Usage with LiteLLM Proxy Server

Here's how to call a XAI model with the LiteLLM Proxy Server

Modify the config.yaml

model_list:
  - model_name: my-model
    litellm_params:
      model: xai/<your-model-name>  # add xai/ prefix to route as XAI provider
      api_key: api-key                 # api key to send your model

Start the proxy

$ litellm --config /path/to/config.yaml

Send Request to LiteLLM Proxy Server

OpenAI Python v1.0.0+
curl

import openai
client = openai.OpenAI(
    api_key="sk-1234",             # pass litellm proxy key, if you're using virtual keys
    base_url="http://0.0.0.0:4000" # litellm-proxy-base url
)

response = client.chat.completions.create(
    model="my-model",
    messages = [
        {
            "role": "user",
            "content": "what llm are you"
        }
    ],
)

print(response)

curl --location 'http://0.0.0.0:4000/chat/completions' \
    --header 'Authorization: Bearer sk-1234' \
    --header 'Content-Type: application/json' \
    --data '{
    "model": "my-model",
    "messages": [
        {
        "role": "user",
        "content": "what llm are you"
        }
    ],
}'

Reasoning Usage

LiteLLM supports reasoning usage for xAI models.

LiteLLM Python SDK
LiteLLM Proxy - OpenAI SDK Usage

reasoning with xai/grok-4.5
import litellm
response = litellm.completion(
    model="xai/grok-4.5",
    messages=[{"role": "user", "content": "What is 101*3?"}],
    reasoning_effort="low",  # low | medium | high
)

print("Reasoning Content:")
print(response.choices[0].message.reasoning_content)

print("\nFinal Response:")
print(response.choices[0].message.content)

print("\nNumber of completion tokens:")
print(response.usage.completion_tokens)

print("\nNumber of reasoning tokens:")
print(response.usage.completion_tokens_details.reasoning_tokens)

reasoning with xai/grok-4.5
import openai
client = openai.OpenAI(
    api_key="sk-1234",             # pass litellm proxy key, if you're using virtual keys
    base_url="http://0.0.0.0:4000" # litellm-proxy-base url
)

response = client.chat.completions.create(
    model="xai/grok-4.5",
    messages=[{"role": "user", "content": "What is 101*3?"}],
    reasoning_effort="low",  # low | medium | high
)

print("Reasoning Content:")
print(response.choices[0].message.reasoning_content)

print("\nFinal Response:")
print(response.choices[0].message.content)

print("\nNumber of completion tokens:")
print(response.usage.completion_tokens)

print("\nNumber of reasoning tokens:")
print(response.usage.completion_tokens_details.reasoning_tokens)

Example Response:

Reasoning Content:
Let me calculate 101 multiplied by 3:
101 * 3 = 303.
I can double-check that: 100 * 3 is 300, and 1 * 3 is 3, so 300 + 3 = 303. Yes, that's correct.

Final Response:
The result of 101 multiplied by 3 is 303.

Number of completion tokens:
14

Number of reasoning tokens:
310

Supported Models​

API Key​

Sample Usage​

Sample Usage - Streaming​

Sample Usage - Vision​

Usage with LiteLLM Proxy Server​

Reasoning Usage​