Skip to content

Can Radeon RX 9070 run MiMo-V2.6-Distill-Qwen-9B?

Comfortable: GGUF Q8_0 needs 10.1 GB at 4K, batch 1; headroom 5.9 GB.

Text input, FP16 KV cache, one card. Unified-memory capacity uses the GPU guide’s usable limit; negative headroom may require host offload. Speed source settings may differ from this 4K estimate.

Measured runs

No measured run for this pair. Sizing is an estimate, not a runtime test.

Quant memory at 4K

QuantTotal GBVerdictWeight sizing
GGUF Q3_K_M4.8ComfortableVerified file
GGUF Q4_K_M6.3ComfortableVerified file
GGUF Q5_K_M7.3ComfortableVerified file
GGUF Q8_010.1ComfortableVerified file

Context capacity · GGUF Q8_0

ContextTotal GBVerdict
4K10.1Comfortable
16K10.5Comfortable
32K11.1Comfortable
128K14.4Tight

Generation speed

GGUF Q8_0: 45–60 tok/s · Derived range

Other model / GPU pairs