rdtand/qwen3.6-27b
Qwen3.6-27B (PrismaQuant)
Throughput
10.6t/s@ 16k
Engine
vLLM
Parameters
27B
Released
2026-04-21
Benchmarked
2026-06-27
Context ladder
Throughput at each benched context window (single measurement).
| Context | KV | Throughput |
|---|---|---|
| @ 16k peak golden | — | 10.3t/s |
Golden profile
rdtand-qwen3-6-27b-eugr
Capabilities
generaldensevllmexperimental
Why we run it
Community PrismaQuant 5.5-bit vLLM quant for Qwen3.6-27B — Spark A/B vs NVFP4/FP8.
Bench notes
golden ?/? @ 10.6 tok/s — fill~5734 — bench-agent-v2 — tool_ok=False