GPU comparison
NVIDIA RTX 5070 Ti (16GB) vs NVIDIA RTX 5090 (32GB) for local LLMs
How the two stack up for running open-source LLMs locally — memory, bandwidth, price, and how many of the 34 tracked models each one can run.
| NVIDIA RTX 5070 Ti (16GB) | NVIDIA RTX 5090 (32GB) | |
|---|---|---|
| Memory | 16.0 GB | 32.0 GB |
| Bandwidth | 896 GB/s | 1792 GB/s |
| Price (approx) | $949 | $4,399 |
| LLMs it runs | 18 of 34 | 25 of 34 |
| Best model it runs | Gemma 4 26B-A4B (MoE) · 161–241 tok/s | Gemma 4 31B · 50–75 tok/s |
The NVIDIA RTX 5090 (32GB) runs 7 more of the tracked models (25 vs 18), thanks to its 32.0 GB of memory. The NVIDIA RTX 5090 (32GB) has more memory bandwidth (1792 GB/s), so it generates tokens faster at the same model and quant.