List your company

ResearchFormerly WizWand

Peak decode throughput

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

SDAR-8B-Chat 32K context

823Throughput (tokens/s)

4

Jun 23, 2026

SDAR-8B-Chat 16K context

1,364Throughput (tokens/s)

4

Jun 23, 2026

SDAR-8B-Chat 8K context

1,484Throughput (tokens/s)

4

Jun 23, 2026

LLaDA mini 16B (32K context) 2.0

2,717Throughput (tokens/s)

4

Jun 23, 2026

LLaDA 16B 16K context 2.0-mini

3,475Throughput (tokens/s)

4

Jun 23, 2026

LLaDA mini 16B 8K context 2.0

3,813Throughput (tokens/s)

4

Jun 23, 2026