Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Long-Context Generation on PG-19 (10K context length)
Loading...
2.92
Throughput Speedup (micro-averaged)
SpecPV-4K
1.3496
1.7573
2.165
2.5727
Dec 2, 2025
Throughput Speedup (micro-averaged)
Draft Accept Length (macro-averaged)
Updated 4d ago
Evaluation Results
Method
Method
Links
Throughput Speedup (micro-averaged)
Draft Accept Length (macro-averaged)
SpecPV-4K
Backbone=LLaMA3.1-8B-I...
2025.12
2.92
3.29
SpecPV-2K
Backbone=LLaMA3.1-8B-I...
2025.12
2.88
3.3
SpecPV-8K
Backbone=LLaMA3.1-8B-I...
2025.12
2.53
3.34
EAGLE3-YARN
Backbone=LLaMA3.1-8B-I...
2025.12
2.49
3.38
TriForce
Backbone=LLaMA3.1-8B-I...
2025.12
1.72
2.62
TokenSwift
Backbone=LLaMA3.1-8B-I...
2025.12
1.41
1.64
Feedback
Search any
task
Search any
task