Can I run DeepSeek-V2-Lite-Chat-Uncensored-Unbiased-Reasoner on a Radeon RX 6500 XT?
Not at these settings. No indexed quantization of DeepSeek-V2-Lite-Chat-Uncensored-Unbiased-Reasoner fits Radeon RX 6500 XT at any context we compute, with q8_0 KV. The smallest shipped quantization is 5.99 GiB in weights alone, against 3.72 GiB usable. CPU offload can still run it, slowly.
Every quantization at every context
| Quant | Weights● | 4K◐ | 8K◐ | 16K◐ | 32K◐ | 64K◐ | 128K◐ |
|---|---|---|---|---|---|---|---|
| Q8_0 | 15.55 GiB | 16.5 | 16.6 | 16.7 | 17.0 | 17.5 | 18.5 |
| Q6_K | 13.10 GiB | 14.1 | 14.1 | 14.3 | 14.5 | 15.0 | 16.0 |
| Q5_K_M | 11.03 GiB | 12.0 | 12.1 | 12.2 | 12.4 | 13.0 | 14.0 |
| Q5_K_S | 10.37 GiB | 11.3 | 11.4 | 11.5 | 11.8 | 12.3 | 13.3 |
| Q4_K_M | 9.65 GiB | 10.6 | 10.7 | 10.8 | 11.1 | 11.6 | 12.6 |
| Q4_K_S | 8.88 GiB | 9.8 | 9.9 | 10.0 | 10.3 | 10.8 | 11.8 |
| IQ4_XS | 8.05 GiB | 9.0 | 9.1 | 9.2 | 9.5 | 10.0 | 11.0 |
| Q3_K_L | 7.88 GiB | 8.8 | 8.9 | 9.0 | 9.3 | 9.8 | 10.8 |
| Q3_K_M | 7.57 GiB | 8.5 | 8.6 | 8.7 | 9.0 | 9.5 | 10.5 |
| Q3_K_S | 6.97 GiB | 7.9 | 8.0 | 8.1 | 8.4 | 8.9 | 9.9 |
| Q2_K | 5.99 GiB | 7.0 | 7.0 | 7.2 | 7.4 | 7.9 | 8.9 |
Figures are GiB of total memory: weights plus KV cache plus compute buffer and backend overhead. Weights and KV are near-exact; the overhead term is modeled. Hover any cell for the breakdown.
Why other calculators disagree
A parameters × bits ÷ 8 estimate ignores two things that dominate at long context. First, the weights themselves are not the nominal rate — quantizations are mixtures, so the real file is consistently larger than the label implies. Second, this model uses latent attention and allocates no V cache at all, so any formula reading num_key_value_heads overstates its cache by more than an order of magnitude.