RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can internlm3 8b instruct run on the RTX 3070?

Yes. At 8K context the best fit is Q4_K_M, using about 6.0 GB of the 7.6 GB usable on the RTX 3070. Estimated speed ceiling: ~84 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •4.0 GB Fits4.2 GB Fits5.4 GB Fits130
3-bit (Q3_K_M) •4.9 GB Fits5.1 GB Fits6.2 GB Fits102
4-bit (Q4_K_M) •5.8 GB Fits6.0 GB Fits7.1 GB Tight84
5-bit (Q5_K_M) •6.7 GB Fits6.9 GB Tight8.0 GB No72
6-bit (Q6_K) •7.6 GB Tight7.8 GB No8.9 GB No—
8-bit (Q8_0) •9.5 GB No9.7 GB No10.9 GB No—
16-bit (F16/BF16)17.2 GB No17.4 GB No18.5 GB No—

• real GGUF file size from internlm/internlm3-8b-instruct-gguf. Speed is a bandwidth ceiling (448 GB/s), not a benchmark. Methodology.

More

internlm3 8b instruct on every GPU and Mac · Every model that runs on the RTX 3070 · Other models that fit 8 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (internlm/internlm3-8b-instruct), NVIDIA specs. Data as of 2026-11-17.