Meta
Llama 4 Scout 109B-A17B
visionMoE
108,6B
Parameters · 17B active
62,9 Go
Q4_K_M file
≥ 64,4 Go
Recommended VRAM
10 485 760
Max context
Not measured yet
No card has run the TokenBench protocol on this model yet. The current protocol covers a selection of models; benchmarking any catalog model from the app is coming soon.
Measure my card →Which cards should it fit on?
Cards on the site whose video memory covers the recommended 64,4 Go (file + 4,000-token context). An estimate, not a measurement. Click a card for the detailed estimate.
Technical sheet
- Published
- 2025-04-02
- Architecture
- 48 layers · 8 KV heads · dimension 128
- Attention
- Standard: every layer keeps the whole context in memory.
- Context cache (f16)
- ≈ 188 MB per 1,000 tokens · 6,0 Go for 32,768 tokens
- GGUF file
- bartowski/meta-llama_Llama-4-Scout-17B-16E-Instruct-GGUF · 2 files Download (62,9 Go)
- Official repo
- meta-llama/Llama-4-Scout-17B-16E-Instruct
Public Hugging Face data (official config.json, GGUF repo). The context cache only counts full-attention layers.