← AI Hashrate

Can RX 9070 16GB run Qwen3-4B-Thinking-2507?

AMD · 16 GB GDDR6 · 640 GB/s bandwidth · Alibaba · 4.0B · model ctx up to 256K

Yes — the RX 9070 16GB (16 GB GDDR6) can run Qwen3-4B-Thinking-2507. At Q4 with the default 8K context it needs ≈ 4.0 GB VRAM (weights 2.2 GB + KV cache 0.8 GB + 1 GB runtime overhead), which fits within 16 GB. Decode at Q4/8K: ≈ 101.8 tok/s (estimated, batch 1). FP16 (≈ 9.8 GB) also fits, at ≈ 28.0 tok/s (estimated). Estimates come from memory-bandwidth math; rows tagged measured override estimates. Relative ranking is more reliable than absolute tok/s. Methodology.

Fit & speed by quant and context

QuantContextVRAM neededFits?tok/s (decode)
Q44K3.6 GBYes101.8 est.
Q48K default4.0 GBYes101.8 est.
Q432K6.4 GBYes101.8 est.
FP164K9.4 GBYes28.0 est.
FP168K default9.8 GBYes28.0 est.
FP1632K12.2 GBYes28.0 est.

Fits = weights + KV(ctx) + 1 GB overhead ≤ 95% of VRAM. Measured anchors are context-agnostic; the fit verdict is recomputed per context. A missing tok/s means the model is far beyond this card (offload-only territory).

VRAM breakdown at 8K context

QuantWeightsKV cacheOverheadTotal neededRX 9070 16GB VRAM
Q42.2 GB0.8 GB1 GB4.0 GB16 GB
FP168.0 GB0.8 GB1 GB9.8 GB16 GB

Other GPUs that run Qwen3-4B-Thinking-2507

All GPUs for Qwen3-4B-Thinking-2507 →

Other models for the RX 9070 16GB

All models on the RX 9070 16GB →