xAI Grok API Pricing Guide 2026: Grok 4.3 and Grok Build 0.1

xAI rebranded its Grok lineup in June 2026. Grok 3 is now Grok 4.3 ($1.25/$2.50) and Grok 3 Mini is now Grok Build 0.1 ($0.30/$0.50). Here's the full breakdown.

xAI Grok Model Lineup

xAI offers two API models, each targeting different use cases and budgets:

Grok 4.3: The Flagship

Grok 4.3 (formerly Grok 3) is xAI's most capable model, designed for complex reasoning, coding, and general-purpose tasks. At $1.25/$2.50 per 1M tokens, it's now priced competitively with GPT-5 ($1.25/$5) and significantly cheaper than Claude Sonnet 4.6 ($3/$15).

Grok Build 0.1: The Budget Option

Grok Build 0.1 (formerly Grok 3 Mini) is xAI's answer to GPT-4o mini and Claude Haiku. At $0.30/$0.50 per 1M tokens, it's significantly cheaper than most competitors while maintaining solid performance for routine tasks.

Use Case Cost Breakdowns

1. Chatbot (10K requests/day)

Assuming 500 input tokens and 300 output tokens per request, running 30 days/month:

Verdict: Grok Build 0.1 is the cheapest at $0.90/month. Grok 4.3 at $4.13 is now 68% cheaper than GPT-4o for chatbot workloads.

2. Code Generation (500 requests/day)

Assuming 2,000 input tokens and 1,500 output tokens per request:

Verdict: Grok 4.3 at $9.38 is now 69% cheaper than GPT-4o for code generation. Grok Build 0.1 at $2.03 is the cheapest option for simpler code tasks.

3. Document Analysis (1,000 requests/day)

Assuming 3,000 input tokens and 500 output tokens per request:

Verdict: Grok 4.3 at $18.75 beats GPT-4o ($37.50) for document analysis. Grok Build 0.1 at $3.45 is the cheapest option for high-volume processing.

When to Choose xAI Grok

Choose Grok 4.3 when:

• You need real-time information about current events or trends
• Your application involves X/Twitter content analysis
• You want fewer content restrictions for creative tasks
• You're already in the xAI/Elon Musk ecosystem

Choose Grok Build 0.1 when:

• You need a budget model with solid general capabilities
• Your use case involves chatbots, classification, or simple Q&A
• You want real-time knowledge at a low price point
• You're testing the xAI ecosystem before committing to Grok 4.3

Consider alternatives when:

• Price is the primary concern: DeepSeek V4 Flash ($0.14/$0.28) is cheaper
• You need the largest context window: Gemini models offer up to 1M tokens
• You need the best code generation: Claude Sonnet 4.6 or GPT-5 may be better
• You need enterprise features: OpenAI and Anthropic have more mature tooling

Cost Optimization Tips for Grok

1. Use Grok Build 0.1 for routine tasks: It's 4x cheaper on input and 5x cheaper on output than Grok 4.3
2. Implement model routing: Route simple queries to Grok Build 0.1, complex ones to Grok 4.3
3. Set max_tokens limits: Grok 4.3 output is $2.50/1M tokens — cap it to avoid runaway costs
4. Batch similar requests: Process multiple documents in a single prompt when possible
5. Monitor usage closely: xAI's usage dashboard may lag — track costs independently

The Bottom Line

xAI Grok occupies a unique niche in the LLM API market. Grok 4.3 at $1.25/$2.50 is now priced competitively with GPT-5 and significantly cheaper than Claude Sonnet 4.6, while offering unique real-time X/Twitter data access. Grok Build 0.1 is a solid budget option, though DeepSeek V4 Flash remains the cheapest overall.

The best strategy: use Grok Build 0.1 for high-volume, low-complexity tasks and reserve Grok 4.3 for tasks where real-time knowledge or creative freedom matters. Combine with other providers for the best cost-to-quality ratio.