Can You Run Qwen3-Coder-Next-abliterated? Every GPU, Tested

Qwen3-Coder-Next-abliterated needs roughly 61GB of VRAM at our tested precision. A model that doesn't fit doesn't run slowly, it doesn't run. Here's every card from our bench, sorted by measured speed.

Weights: bartowski/huihui-ai_Qwen3-Coder-Next-abliterated-GGUF

GPUs that run Qwen3-Coder-Next-abliterated (6)

GPUSpeedVRAMBasis
NVIDIA RTX PRO 6000 Blackwell Workstation Edition210.47 tok/s96GB✓ Measured
NVIDIA B300199.12 tok/s288GB✓ Measured
NVIDIA H200197.14 tok/s141GB✓ Measured
NVIDIA H100 80GB HBM3190.32 tok/s80GB✓ Measured
NVIDIA B200186.57 tok/s192GB✓ Measured
NVIDIA A100 80GB SXM4119.56 tok/s80GB✓ Measured

GPUs it won't fit on (5)

These cards' VRAM is below the requirement at tested precision, published as data, not hidden.

GPUVRAM
NVIDIA L40S48GB
NVIDIA A100 40GB SXM440GB
NVIDIA A10G24GB
NVIDIA L424GB
NVIDIA T416GB

Quick answers

How much VRAM does Qwen3-Coder-Next-abliterated need?

Qwen3-Coder-Next-abliterated needs roughly 61GB of VRAM at our tested precision. 6 of the GPUs on our bench run it; 5 don't have the VRAM for it.

What is the fastest GPU for Qwen3-Coder-Next-abliterated?

NVIDIA RTX PRO 6000 Blackwell Workstation Edition is the fastest card we've measured running Qwen3-Coder-Next-abliterated, at 210.47 tok/s.

What is the cheapest GPU that can run Qwen3-Coder-Next-abliterated?

NVIDIA RTX PRO 6000 Blackwell Workstation Edition ($8,565 MSRP) is the cheapest card on our bench that runs Qwen3-Coder-Next-abliterated, at 210.47 tok/s.

← All models · AI GPU rankings · How we benchmark