All models
Meta
Llama 4 Scout fastruns locally
meta/llama-4-scout
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...
Input / 1M
$0.10
USD per million tokens
Output / 1M
$0.30
USD per million tokens
Est. cost / token
$0.00000015
$0.15 per 1M, 3:1 mix
Context window
328K
max output 16K
Major features
VisionOpen weightsJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding3
- reasoning2
- math2
- summarization4
- extraction4
- classification4
- chat3
- creative3
- translation4
- vision3
- long context3
- agentic3
Run it locally
Will it run on my machine? →ollama run llama4:scout| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 51.7 GB |
| Q3_K_M | fair | 59.8 GB |
| Q4_K_Mrecommended | good | 73.8 GB |
| Q5_K_M | very good | 85.2 GB |
| Q6_K | excellent | 97.7 GB |
| Q8_0 | near-lossless | 125 GB |
| F16 | lossless | 233 GB |
Estimates. Mixture-of-experts: memory covers all parameters; only the active ones are used per token, so it runs faster than its size suggests.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.10 | $0.30 | — | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.10
- Output
- $0.30
- Blended (3:1)
- $0.15
Details
- Company
- Meta
- Version
- 4
- Family
- Llama
- Released
- 2025-04-05
- Parameters
- 109B (17B active)
- Max output
- 16K
- Open weights
- Yes
- Inputs
- image, text
- Last synced
- 46 min ago