Huihui-Qwen3.8-27B-abliterated

Qwen family · 28.0B params

Evidence: 1 sourced run · 1 independent source · first recorded Aug 18, 2026. What these labels mean.

What the Huihui-Qwen3.8-27B-abliterated numbers mean

The fastest verified result on our board is 117.7 output tok/s on an RTX 5090 at NVFP4 with a 172800-token context. That is about 8.5 ms per generated token — 500 tokens takes roughly 4.2 seconds. Prompt processing on the same record runs at 7466.2 tok/s, so a 2,000-token prompt is read in about 0.3 seconds.

How much VRAM does Huihui-Qwen3.8-27B-abliterated need?

Huihui-Qwen3.8-27B-abliterated has 28.0B parameters. These estimates use the same formula as the VRAM calculator: weights, plus about 15% for a moderate-context KV cache and runtime overhead, plus a small engine floor.

Long contexts and large batches raise the real number — read the smallest card as "possible with a reduced context". Measured throughput is on the benchmark board.

Models with a similar footprint

Ranked by distance from Huihui-Qwen3.8-27B-abliterated's parameter count. Full benchmark board.

All benchmarks Methodology