Pricing Guide · Meta

Llama 4 Scout API Pricing

Detailed token costs, context window limits, benchmark performance, and caching structures for Llama 4 Scout.

Last reviewed: June 6, 2026 Sourced from official docs

Llama 4 Scout Cost Breakdown

Dimension	Value / Cost
Input Tokens Cost / 1M	Self-hosted
Output Tokens Cost / 1M	Self-hosted
Cached Input Tokens Cost / 1M	—
Max Context Window	10,000,000 tokens
SWE-bench Verified Score	56.0%
GPQA Diamond Score	57.2%

Best suited for:

Ultra-long-context tasks (10M token window)
Fast self-hosted inference
Free multimodal at scale

← Back to all pricing guides