---
id: "meta-llama/llama-4-scout"
also: ["meta-llama/llama-4-scout-17b-16e-instruct", "meta-llama/Llama-4-Scout-17B-16E-Instruct", "hf/meta-llama/Llama-4-Scout-17B-16E-Instruct", "vercel/meta/llama-4-scout"]
cheapest_list_usd_per_m: "0.08 in, 0.24 out, at deepinfra/fp8/flex on openrouter"
cheapest_effective_usd_per_m: "0.0797 in, 0.2301 out, at deepinfra/fp8/flex on openrouter"
context: 327680
max_output: 16384
input: ["text", "image"]
output: ["text"]
released: "2025-04-05"
api: ["/api/v1/chat/completions", "/api/v1/completions", "/api/v1/messages", "/api/v1/responses"]
params: ["frequency_penalty", "include_reasoning", "logit_bias", "max_tokens", "min_p", "presence_penalty", "reasoning", "repetition_penalty", "response_format", "seed", "stop", "structured_outputs", "temperature", "tool_choice", "tools", "top_k", "top_p"]
endpoints: 12
gateways: "openrouter (6), hf (3), vercel (3)"
no_training: 9
trains: 0
retains_prompts: 1
coding: 8.2000
agentic: 0.5000
---
# Meta: Llama 4 Scout

Llama 4 Scout is the best multimodal model in the world in its class and is more powerful than our Llama 3 models, while fitting in a single H100 GPU. Additionally, Llama 4 Scout supports an industry-leading context window of up to 10M tokens.