RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can Ling 3.0 tiny run on the H100 SXM?

Yes. At 8K context the best fit is F16, using about 15.6 GB of the 76.0 GB usable on the H100 SXM. Estimated speed ceiling: ~1211 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •3.6 GB Fits3.7 GB Fits4.3 GB Fits6.9 GB Fits6379
3-bit (Q3_K_M) •4.3 GB Fits4.4 GB Fits5.1 GB Fits7.6 GB Fits5055
4-bit (Q4_K_M) •5.4 GB Fits5.5 GB Fits6.1 GB Fits8.7 GB Fits3892
5-bit (Q5_K_M) •6.1 GB Fits6.2 GB Fits6.9 GB Fits9.4 GB Fits3348
6-bit (Q6_K) •7.2 GB Fits7.3 GB Fits7.9 GB Fits10.4 GB Fits2798
8-bit (Q8_0) •8.6 GB Fits8.7 GB Fits9.4 GB Fits11.9 GB Fits2276
16-bit (F16/BF16) •15.5 GB Fits15.6 GB Fits16.3 GB Fits18.8 GB Fits1211

• real GGUF file size from bartowski/Ling-3.0-tiny-GGUF. Speed is a bandwidth ceiling (3350 GB/s), not a benchmark. Methodology.

More

Ling 3.0 tiny on every GPU and Mac · Every model that runs on the H100 SXM · Other models that fit 64 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (inclusionAI/Ling-3.0-tiny), NVIDIA specs. Data as of 2026-11-17.