Llama 4 Scout
by Meta
Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimo
Models
Llama 4 Scout
Added 10 July 2026
Overview
Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input...
Best for
Best for
Open-weight long-context coding on a budget
Use cases
- repo Q&A
- self-host experiments
- batch coding
How to use / API access
Call via OpenRouter with model id `meta-llama/llama-4-scout`.
- openrouter ·
meta-llama/llama-4-scoutDocs
OpenRouter id: meta-llama/llama-4-scout
Benchmarks
- Intelligence Index (artificial-analysis) : 10 index
- Coding Index (artificial-analysis) : 8.2 index
- Agentic Index (artificial-analysis) : 1.1 index
- Design Arena codecategories (other) : 833 elo · #108
- Design Arena dataviz (other) : 937 elo · #98
- Design Arena gamedev (other) : 838 elo · #106
- Design Arena uicomponent (other) : 821 elo · #101
Enterprise DNA angle
Omni helps teams prove Scout quality on real jobs before committing infra.
Notes
Llama 4 Scout is listed in the EDNA Models directory from the OpenRouter catalogue.
Pros
- Open-weights economics with long context
- Good fit for cost-controlled coding agents
Cons
- Hosted quality varies by endpoint
- May need eval gates vs closed frontier models