Llama 4 Scout API Prices Drop 49%: Input Now $0.10, Output $0.30 per 1M Tokens

What Changed

Meta's Llama 4 Scout has received a substantial price cut across both input and output tokens. Input pricing dropped from $0.18 to $0.10 per 1M tokens, and output pricing fell from $0.59 to $0.30 per 1M tokens — a -49% reduction across the board.

Does This Move the Needle?

For output-heavy workloads — think long-form generation, summarization pipelines, or agentic tasks that produce verbose responses — the drop from $0.59 to $0.30 on output tokens is the more meaningful number. Output tokens are almost always the dominant cost driver, so cutting them nearly in half has a real compounding effect at scale.

For a workload burning 100M output tokens a month, that's a swing from $59 to $30 — meaningful budget relief without touching a single line of code.

Input-side savings are more modest in absolute terms, but the move from $0.18 to $0.10 still improves the economics of retrieval-augmented generation setups where large context windows get stuffed with documents.

Who Should Care

• High-volume API users already on Scout will see immediate savings with no migration cost.
• Teams sitting on the fence between Scout and a pricier alternative may find the calculus has shifted.
• Budget-conscious experimenters now have less reason to throttle usage during development.

Check the Llama 4 Scout — live specs & price history to compare it against alternatives currently listed on the site before committing to a workload.

Bottom Line

This is a clean, no-asterisks price cut. No capability changes, no tier restrictions mentioned — just lower numbers. If you're evaluating long-context or high-throughput use cases, Scout's updated pricing makes it worth a fresh look.

Ezra, Scout AI Team