RigBench data 2026-10-22

Home › Models › DeepSeek R1 Distill Qwen 1.5B

DeepSeek R1 Distill Qwen 1.5B: VRAM needs and speed by GPU

At Q4_K_M, DeepSeek R1 Distill Qwen 1.5B needs about 2.0 GB of VRAM. We have 1 measured speed results for it.

1.78B parameters · GGUF files on Hugging Face

Measured speeds

ChipGeneration tok/sPrompt tok/sQuantBackendTestSource
Apple M4 Max (40-core GPU)333–MLX-4bitMLX short-chat decode (warm median)john-rocky apple-silicon-llm-bench (MIT) details

Memory needed per quantization

QuantFile sizeNeeds (estimate)Smallest card it fits
IQ2_M0.7 GB1.5 GB8 GB cards
Q2_K0.8 GB1.6 GB8 GB cards
IQ3_XS0.8 GB1.7 GB8 GB cards
Q3_K_S0.9 GB1.7 GB8 GB cards
IQ3_M0.9 GB1.7 GB8 GB cards
Q3_K_M0.9 GB1.8 GB8 GB cards
Q3_K_L1.0 GB1.8 GB8 GB cards
Q2_K_L1.0 GB1.8 GB8 GB cards
IQ4_XS1.0 GB1.9 GB8 GB cards
IQ4_NL1.1 GB1.9 GB8 GB cards
Q4_01.1 GB1.9 GB8 GB cards
Q4_K_S1.1 GB1.9 GB8 GB cards
Q4_K_M1.1 GB2.0 GB8 GB cards
Q4_11.2 GB2.0 GB8 GB cards
Q3_K_XL1.2 GB2.0 GB8 GB cards
Q5_K_S1.3 GB2.1 GB8 GB cards
Q5_K_M1.3 GB2.1 GB8 GB cards
Q4_K_L1.3 GB2.1 GB8 GB cards
Q5_K_L1.4 GB2.3 GB8 GB cards
Q6_K1.5 GB2.3 GB8 GB cards
Q6_K_L1.6 GB2.4 GB8 GB cards
Q8_01.9 GB2.7 GB8 GB cards
F163.6 GB4.4 GB8 GB cards

Expected speed on popular hardware

HardwareBest quant that fitsExpected tok/s
NVIDIA RTX 3060Q8_0163 estimate
NVIDIA RTX 4060 TiQ8_0129 estimate
NVIDIA RTX 3090Q8_0346 estimate
NVIDIA RTX 4090Q8_0460 estimate
NVIDIA RTX 5090Q8_0606 estimate
AMD Radeon RX 7900 XTXQ8_0385 estimate
Apple M4 Pro (20-core GPU) (64 GB)Q8_0122 estimate
Apple M3 Max (40-core GPU) (64 GB)Q8_0178 estimate
AMD Ryzen AI Max+ 395 (Radeon 8060S)Q8_0110 estimate

Check your own card

Too big for your card?

Run DeepSeek R1 Distill Qwen 1.5B on a rented 24-80 GB GPU by the hour.

Rent a bigger GPU on RunPod

Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.