RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can deepseek coder 6.7b instruct run on the RTX 3070?

Yes. At 8K context the best fit is Q3_K_M, using about 7.5 GB of the 7.6 GB usable on the RTX 3070. Estimated speed ceiling: ~136 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •5.1 GB Fits7.1 GB Tight158
3-bit (Q3_K_M) •5.5 GB Fits7.5 GB Tight136
4-bit (Q4_K_M) •6.3 GB Fits8.3 GB No—
5-bit (Q5_K_M) •6.9 GB Tight8.9 GB No—
6-bit (Q6_K) •7.6 GB No9.6 GB No—
8-bit (Q8_0) •9.1 GB No11.1 GB No—
16-bit (F16/BF16)15.0 GB No17.0 GB No—

• real GGUF file size from TheBloke/deepseek-coder-6.7B-instruct-GGUF. Speed is a bandwidth ceiling (448 GB/s), not a benchmark. Methodology.

More

deepseek coder 6.7b 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 (deepseek-ai/deepseek-coder-6.7b-instruct), NVIDIA specs. Data as of 2026-11-17.