Qwen235B params22B active (MoE)128k contextpopular
Qwen 3 235B A22B locally
Qwen 3 235B A22B is an open-weight LLM from the Qwen family with 235B parameters. Main use: chat, RAG and general assistance. Detected minimum hardware: NVIDIA B100 (192 GB).
Technical facts
Parameters235B
Max context128k
Q4_K_M147.7 GB
Q5_K_M180.6 GB
Q8262.6 GB
FP16525.3 GB
FamilyQwen
Last sync2026-05-12
Available quantizations
GGUF weightsQ4_K_M
147.7GB
Acceptable. Good compromise when VRAM is limited.
Q5_K_M
180.6GB
Good quality. Sweet spot for size and precision.
Q8
262.6GB
Near-FP16 quality. Comfortable for production.
FP16
525.3GB
Reference precision. Maximum quality, doubled VRAM.
Compatible GPUs
10 single-GPUGPUs that can run Qwen 3 235B A22B on a single card, ranked by VRAM headroom.
NVIDIA B100
NVIDIA192 GB · Blackwell DC
147.7 / 192 GBcomfortable · Q4
NVIDIA B200
NVIDIA192 GB · Blackwell DC
147.7 / 192 GBcomfortable · Q4
NVIDIA GB200 (per GPU)
NVIDIA192 GB · Blackwell DC
147.7 / 192 GBcomfortable · Q4
Instinct MI300X
AMD192 GB · Instinct CDNA 3+
147.7 / 192 GBcomfortable · Q4
Mac Studio M2 Ultra (192GB)
Apple192 GB · Mac Studio
147.7 / 192 GBcomfortable · Q4
Mac Studio M3 Ultra (192GB)
Apple192 GB · Mac Studio
147.7 / 192 GBcomfortable · Q4
Mac Pro M2 Ultra (192GB)
Apple192 GB · Mac Pro
147.7 / 192 GBcomfortable · Q4
Instinct MI325X
AMD256 GB · Instinct CDNA 3+
180.6 / 256 GBcomfortable · Q5
Mac Studio M3 Ultra (256GB)
Apple256 GB · Mac Studio
180.6 / 256 GBcomfortable · Q5
Mac Studio M3 Ultra (512GB)
Apple512 GB · Mac Studio
262.6 / 512 GBcomfortable · Q8
Similar models
VRAM estimates: parameters x bits/8 plus margin. Real performance varies by engine, context length and batch size.
sync: 2026-05-12