selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

tencent logo

Can the RTX 4070 run Qwen3.8-27B-DSpine-b16?

Qwen3.8-27B-DSpine-b16·2B·GeForce RTX 4070·12 GB VRAM

Saved in this browser, used on every model page.
Yes

Yes. The RTX 4070 runs Qwen3.8-27B-DSpine-b16 fully in VRAM at FP16 / BF16 (about 4.6 GB of 12 GB), at roughly 68 tokens/s.

best quant
FP16 / BF16
needs
4.6 GB
usable
12 GB
est. speed
~68 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the Qwen3.8-27B-DSpine-b16 page →

Every quantisation of Qwen3.8-27B-DSpine-b16 on the RTX 4070

QuantQualityMemoryFitSpeed
Q2_K Heavy loss1.3 GBfits~416 tok/s
Q3_K_M Noticeable loss1.6 GBfits~280 tok/s
Q4_K_M Small loss1.9 GBfits~226 tok/s
Q5_K_M Near-lossless2.1 GBfits~192 tok/s
Q6_K Near-lossless2.3 GBfits~166 tok/s
Q8_0 Lossless2.8 GBfits~129 tok/s
FP16 / BF16 pickLossless4.6 GBfits~68 tok/s

Speed is a bandwidth-based estimate for generating text (504 GB/s); prompt processing and long contexts are slower.

Measured speed on the RTX 4070

No community reports for this pair yet.

Report your own speed →

Chat models that fit the RTX 4070 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
2BQ8_0≈3.2 GBFits in memory~128 tok/sCommercial OK
LiquidAI logoLFM3-1.7BLiquidAI
1.7BQ8_0≈2.7 GBFits in memory~151 tok/sConditions
1.5BQ8_0≈2.6 GBFits in memory~169 tok/sCommercial OK
2.5BQ8_0≈3.5 GBFits in memory~104 tok/sCommercial OK
3.2BQ8_0≈4.7 GBFits in memory~81 tok/sConditions
2.6BQ8_0≈4.0 GBFits in memory~100 tok/sConditions

Other GPUs that run Qwen3.8-27B-DSpine-b16

Frequently asked questions

Can the RTX 4070 run Qwen3.8-27B-DSpine-b16?
Yes. The RTX 4070 runs Qwen3.8-27B-DSpine-b16 fully in VRAM at FP16 / BF16 (about 4.6 GB of 12 GB), at roughly 68 tokens/s.
How fast does Qwen3.8-27B-DSpine-b16 run on the RTX 4070?
An estimate from memory bandwidth gives about 68 tokens/s at FP16 / BF16.
Which quantisation of Qwen3.8-27B-DSpine-b16 should I use on the RTX 4070?
FP16 / BF16: the largest one that fits. Expected quality: lossless compared with the full-precision weights.

Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.