Loading...
Loading...
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
$0.08
per 1M tokens
Output
$0.30
per 1M tokens
Blended
$0.14
per 1M tokens
Cheaper than 79% of models. Median price is $0.56/1M tokens.
Daily
$0.14
Monthly
$4.05
123
tokens/sec
Faster than 67% of models
0.56
seconds
Faster than 77% of models
0.56
seconds
Faster than 84% of models
Market Median
86 tok/s
43% faster
Median TTFT
1.07s
47% faster
Throughput/Dollar
912
tok/s per $/1M
Speed Comparison
Context Window
328K
tokens
Larger than 79% of models
Max Output
16K
tokens
5% of context
Quality Index
13.5
359th of 507
Top 71%
Coding Index
6.7
359th of 417
Top 87%
Math Index
14.0
221st of 269
Top 83%
Price/1M
$0.14
132nd cheapest
76% below median
Top 21%
Speed
123 tok/s
Top 33%
TTFT
0.56s
Context Window
328K
87th largest
Top 21%