Back to Developer Zone

xAI Models

Explore all 3 models from xAI with detailed pricing, pros & cons, and developer recommendations.

3
Models
$0.200
Lowest Input
2M
Max Context
2
Quality Tiers

Quick Recommendations

Best Value: Grok 4.1 Fast ($0.200/1M)
Best Quality: Grok 4.3

Grok 4.3

Flagship

General-purpose, max capability

Official Pricing

When to use: Best xAI model for general-purpose applications requiring real-time knowledge access.

Upgrade Highlights

  • 1M context window — 8x increase over Grok 3's 128K
  • Price: $1.25/$2.50 — 60% cheaper than Grok 3 ($3/$15)
  • Real-time web + X (Twitter) search — unique live knowledge access
  • Context caching: $0.20/M — 84% savings for repeated prefixes
  • 32K max output — 4x Grok 3's 8K for long-form generation
Input Price
$1.25
per 1M tokens
Output Price
$2.50
per 1M tokens
Cached Input
$0.200
per 1M tokens
Batch Input
per 1M tokens
Context Window: 1M
Max Output: 32,000 tokens
Knowledge Cutoff: 2025-06
VisionFunction CallingFine-tuningJSON Mode

Pros

  • 1M context window at $1.25/M input
  • Most intelligent xAI model
  • Real-time web and X search

Cons

  • No batch API
  • No fine-tuning
  • Newer ecosystem

Performance

Output Speed~65 tok/s
Rate Limit3,000 RPM

Multimodal

Image InputImage OutputAudio InputAudio Output

Benchmarks

MMLU
87.5%
SWE-bench Verified
62.0%
GPQA
73.0%

Grok 4.1 Fast

Mid-tier

Cost-optimized production workloads

Official Pricing

When to use: Best for latency-sensitive production apps and long-document processing on a budget.

Upgrade Highlights

  • 2M context window — largest among all mid-tier models
  • $0.20/M input — 6x cheaper than Grok 4.3 for high-volume tasks
  • Ultra-low latency — optimized for sub-500ms response times
  • 16K max output — 2x Grok 3's 8K for longer generations
  • Context caching: $0.05/M — 75% savings for repeated prefixes
Input Price
$0.200
per 1M tokens
Output Price
$0.500
per 1M tokens
Cached Input
$0.050
per 1M tokens
Batch Input
per 1M tokens
Context Window: 2M
Max Output: 16,000 tokens
Knowledge Cutoff: 2025-04
VisionFunction CallingFine-tuningJSON Mode

Pros

  • 2M context window — largest among mid-tier
  • Extremely fast response times
  • One of lowest rates for frontier APIs

Cons

  • Lower quality than Grok 4.3
  • No batch API
  • No fine-tuning

Performance

Output Speed~110 tok/s
Rate Limit8,000 RPM

Multimodal

Image InputImage OutputAudio InputAudio Output

Benchmarks

MMLU
83.0%
HumanEval
80.0%

Grok 3

Flagship

Real-time info, analysis

Official Pricing

When to use: When real-time X/Twitter data or very current knowledge is essential.

Upgrade Highlights

  • Real-time X/Twitter integration — unique live social media knowledge
  • Knowledge cutoff: 2025-02 — one of the most current available
  • Vision + function calling — full multimodal capability at launch
  • No caching or batch API — upgrade to Grok 4.3 for caching support
  • 128K context — superseded by Grok 4.3's 1M for long-document tasks
Input Price
$3.00
per 1M tokens
Output Price
$15.00
per 1M tokens
Cached Input
per 1M tokens
Batch Input
per 1M tokens
Context Window: 131K
Max Output: 8,192 tokens
Knowledge Cutoff: 2025-02
VisionFunction CallingFine-tuningJSON Mode

Pros

  • Very recent knowledge cutoff
  • Real-time X/Twitter integration
  • Vision + function calling

Cons

  • Expensive ($3/$15)
  • No caching or batch
  • No fine-tuning

Performance

Output Speed~55 tok/s
Rate Limit2,000 RPM

Multimodal

Image InputImage OutputAudio InputAudio Output

Benchmarks

MMLU
85.0%
GPQA
68.0%
MATH
70.0%

Side-by-Side Comparison

ModelTierInputOutputContext
Grok 4.3Flagship$1.25$2.501M
Grok 4.1 FastMid-tier$0.200$0.5002M
Grok 3Flagship$3.00$15.00131K