Llama 4 Scout

Pricing updated daily · Last refreshed Oct 29, 2026

Meta model with a 1.31M-token context window, priced at $0.10 per million input tokens and $0.30 per million output tokens.

At a glance

Provider
Meta
Context window
1,310,720 tokens
Max output
16,384 tokens
Input price
$0.10 / 1M tokens
Output price
$0.30 / 1M tokens
Input types
text, image
Function calling
Yes
Released
April 5, 2025

What it costs

Illustrative request sizes, not measurements of any real product. For your own volumes, use the cost estimator.

Example costs for Llama 4 Scout
WorkloadTokens in / outPer requestPer 1,000 requests
Short chat reply1,000 / 500$0.000250$0.2500
Question over retrieved documents8,000 / 500$0.000950$0.9500
Summarise a long document50,000 / 1,500$0.0055$5.45

How it compares

  • Input price ($0.10 per 1M tokens) is the 3rd lowest of 30 models in the catalogue.
  • Output price ($0.30 per 1M tokens) is the 5th lowest of 30.
  • Context window (1.31M tokens) is the largest of 30.
  • Maximum output (16K tokens) is the 23rd largest of 30.
  • On blended price it is 3% cheaper than the median Meta model in the catalogue.

These are rankings by listed price and size within this catalogue (blended price weights input 3:1 against output), not a measure of output quality.

More from Meta

Other Meta models
ModelProviderContextInput $/1MOutput $/1M
Llama Guard 4 12BMeta164K$0.18$0.18
Llama 4 MaverickMeta524K$0.188$0.652
Llama 3.3 70B InstructMeta131K$0.10$0.32
Llama 3.2 1B InstructMeta60K$0.027$0.201

Try it with your own text

Source: OpenRouter’s public model catalogue (meta-llama/llama-4-scout), read once a day. These are the prices OpenRouter charges; a provider’s own list price can differ, so check the provider before committing to a budget.

llmtools is a vendor-neutral toolkit for working with large language models — token counting, cost estimation and model comparison, computed in your browser. OpenAI token counts are exact, using the same encoding the model does (o200k_base for GPT-4o, cl100k_base for GPT-4/3.5); Claude and Gemini counts are labelled approximations, since neither publishes a tokenizer that runs client-side. Model pricing is read from OpenRouter’s public catalogue once a day. Nothing you type is ever transmitted anywhere.