choosorai
Meta · Current · OSS

Llama 4 Scout

Open-source Meta model with mixture-of-experts architecture (17B active parameters, 109B total) and exceptional context window up to 10M tokens.

Reasoning
Speed
Context window
10M
Input
Text input supportedImage input supportedAudio input not supportedVideo input not supported
Output
Text output supportedImage output not supportedAudio output not supportedVideo output not supported

Pricing

per 1M tokens

Input price
$0.08
Cache read price
Cache write price
Output price
$0.30

Preise sind Median-Aggregat über mehrere Hosting-Anbieter (OpenRouter). Günstigster Anbieter Groq: $0.11 / $0.34 per 1M.

Capabilities

  • Function calling
  • Vision
  • Structured output
  • Extended reasoning
  • Prompt caching
  • Batch API
  • Streaming

Limits & knowledge

Context window
10M
Max output
Knowledge cutoff
Aug 2024
Release date
Apr 5, 2025
Family
Llama 4

Source & last check

Source: https://openrouter.ai/models/meta-llama/llama-4-scout · Verified on: May 25, 2026