RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can Kimi Linear 48B A3B Instruct run on the 2x RTX 5090?

Yes. At 8K context the best fit is Q8_0, using about 49.6 GB of the 60.8 GB usable on the 2x RTX 5090. Estimated speed ceiling: ~562 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •17.2 GB Fits17.3 GB Fits18.0 GB Fits20.8 GB Fits1674
3-bit (Q3_K_M) •22.0 GB Fits22.1 GB Fits22.8 GB Fits25.6 GB Fits1294
4-bit (Q4_K_M) •28.8 GB Fits28.9 GB Fits29.7 GB Fits32.5 GB Fits976
5-bit (Q5_K_M) •33.5 GB Fits33.6 GB Fits34.3 GB Fits37.2 GB Fits836
6-bit (Q6_K) •38.5 GB Fits38.6 GB Fits39.3 GB Fits42.2 GB Fits725
8-bit (Q8_0) •49.5 GB Fits49.6 GB Fits50.3 GB Fits53.2 GB Fits562
16-bit (F16/BF16) •92.4 GB No92.5 GB No93.2 GB No96.0 GB No—

• real GGUF file size from bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF. Speed is a bandwidth ceiling (1792 GB/s), not a benchmark. Methodology.

More

Kimi Linear 48B A3B Instruct on every GPU and Mac · Every model that runs on the 2x RTX 5090 · Try other settings in the calculator

Data: Hugging Face Hub (moonshotai/Kimi-Linear-48B-A3B-Instruct), NVIDIA specs. Data as of 2026-11-17.