Agent skill
xai-models
xAI Grok model selection and capabilities guide. Use when choosing the right Grok model for your task, comparing model features, or optimizing costs.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/xai-models
SKILL.md
xAI Grok Models Guide
Complete guide to selecting the right Grok model for your use case, with pricing and capability comparisons.
Model Quick Reference
| Model | Best For | Input $/1M | Output $/1M | Context |
|---|---|---|---|---|
grok-4-1-fast |
Tool calling, agents | $0.20 | $0.50 | 2M |
grok-4 |
Complex reasoning | $3.00 | $15.00 | 256K |
grok-3-fast |
General tasks | $0.20 | $0.50 | 131K |
grok-3-mini |
Lightweight tasks | $0.30 | $0.50 | 131K |
grok-2-vision |
Image analysis | $2.00 | $10.00 | 32K |
Model Selection Decision Tree
What's your primary need?
│
├─► Tool calling / Agent workflows
│ └─► grok-4-1-fast ($0.20/$0.50)
│
├─► Complex reasoning / Analysis
│ └─► grok-4 ($3.00/$15.00)
│
├─► General chat / Simple tasks
│ └─► grok-3-fast ($0.20/$0.50)
│
├─► High volume / Cost sensitive
│ └─► grok-3-mini ($0.30/$0.50)
│
└─► Image/Vision tasks
└─► grok-2-vision ($2.00/$10.00)
Detailed Model Profiles
grok-4-1-fast (Recommended for Most Uses)
Best for: Tool calling, agentic workflows, real-time search
# Best choice for X search and sentiment analysis
response = client.chat.completions.create(
model="grok-4-1-fast",
messages=[{"role": "user", "content": "Search X for AAPL sentiment"}]
)
Features:
- 2 million token context window
- Optimized for tool calling
- Fast response times
- Best price/performance ratio
Variants:
grok-4-1-fast-reasoning- Maximum intelligencegrok-4-1-fast-non-reasoning- Instant responses
grok-4
Best for: Deep analysis, complex reasoning, research
# Use for complex multi-step analysis
response = client.chat.completions.create(
model="grok-4",
messages=[{"role": "user", "content": "Analyze market trends..."}]
)
Features:
- Highest reasoning capability
- Best for complex tasks
- 256K context window
grok-3-fast
Best for: General purpose, balanced performance
# Good default choice for most tasks
response = client.chat.completions.create(
model="grok-3-fast",
messages=[{"role": "user", "content": "Summarize this..."}]
)
Features:
- Fast responses
- 131K context
- Good balance of speed/quality
grok-3-mini
Best for: High-volume, cost-sensitive applications
# Use for bulk processing
response = client.chat.completions.create(
model="grok-3-mini",
messages=[{"role": "user", "content": "Classify: ..."}]
)
Features:
- Lowest latency
- Most cost-effective
- Good for simple tasks
grok-2-vision
Best for: Image analysis, charts, screenshots
import base64
# Encode image
with open("chart.png", "rb") as f:
image_data = base64.b64encode(f.read()).decode()
response = client.chat.completions.create(
model="grok-2-vision",
messages=[{
"role": "user",
"content": [
{"type": "text", "text": "Analyze this chart"},
{"type": "image_url", "image_url": {"url": f"data:image/png;base64,{image_data}"}}
]
}]
)
Cost Optimization Strategies
1. Use the Right Model
# For filtering/classification - use mini
filter_response = client.chat.completions.create(
model="grok-3-mini",
messages=[{"role": "user", "content": f"Is this relevant? {text}"}]
)
# For analysis - use fast
if is_relevant:
analysis = client.chat.completions.create(
model="grok-4-1-fast",
messages=[{"role": "user", "content": f"Analyze: {text}"}]
)
2. Leverage Caching
Cached input tokens are 75% cheaper:
- Regular: $0.20/1M
- Cached: $0.05/1M
3. Batch Similar Requests
# Instead of 10 separate calls, batch them
texts = ["text1", "text2", "text3"]
batch_prompt = "Analyze these texts:\n" + "\n".join(texts)
response = client.chat.completions.create(
model="grok-3-fast",
messages=[{"role": "user", "content": batch_prompt}]
)
Tool Calling Costs
| Tool | Cost per 1,000 calls |
|---|---|
| X Search | $5.00 |
| Web Search | $5.00 |
| Code Execution | $5.00 |
| Document Search | $2.50 |
Context Window Comparison
| Model | Context | Pages of Text | Hours of Audio |
|---|---|---|---|
| grok-4-1-fast | 2M | ~6,000 | ~50 |
| grok-4 | 256K | ~800 | ~6 |
| grok-3-fast | 131K | ~400 | ~3 |
| grok-2-vision | 32K | ~100 | ~1 |
Model Capabilities Matrix
| Capability | 4.1 Fast | 4 | 3 Fast | 3 Mini | 2 Vision |
|---|---|---|---|---|---|
| Tool Calling | ⭐⭐⭐ | ⭐⭐ | ⭐ | ⭐ | ❌ |
| Reasoning | ⭐⭐ | ⭐⭐⭐ | ⭐⭐ | ⭐ | ⭐⭐ |
| Speed | ⭐⭐⭐ | ⭐ | ⭐⭐⭐ | ⭐⭐⭐ | ⭐⭐ |
| Cost | ⭐⭐⭐ | ⭐ | ⭐⭐⭐ | ⭐⭐⭐ | ⭐⭐ |
| Vision | ❌ | ❌ | ❌ | ❌ | ⭐⭐⭐ |
| X Search | ⭐⭐⭐ | ⭐⭐ | ⭐⭐ | ⭐ | ❌ |
Recommended Configurations
Financial Sentiment Pipeline
MODELS = {
"filter": "grok-3-mini", # Fast filtering
"analyze": "grok-4-1-fast", # Tool calling + analysis
"deep": "grok-4" # Complex reasoning (rare)
}
High-Volume Processing
MODELS = {
"bulk": "grok-3-mini",
"quality_check": "grok-3-fast"
}
Research & Analysis
MODELS = {
"search": "grok-4-1-fast",
"analyze": "grok-4",
"summarize": "grok-3-fast"
}
API Usage Example
import os
from openai import OpenAI
client = OpenAI(
api_key=os.getenv("XAI_API_KEY"),
base_url="https://api.x.ai/v1"
)
# List available models
models = client.models.list()
for model in models.data:
print(f"{model.id}")
# Use specific model
response = client.chat.completions.create(
model="grok-4-1-fast",
messages=[{"role": "user", "content": "Hello!"}],
max_tokens=100
)
Related Skills
xai-auth- Authentication setupxai-agent-tools- Tool callingxai-sentiment- Sentiment analysis
References
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
majiayu000/claude-skill-registry
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
majiayu000/claude-skill-registry
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
majiayu000/claude-skill-registry
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
majiayu000/claude-skill-registry
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
majiayu000/claude-skill-registry
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
majiayu000/claude-skill-registry
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?