RTX 5070 Ti vs RTX 5080 for local LLMs
For local LLMs these two are close; let price and the rest of your build decide.
- Both run the same number of the 60 most-downloaded models at 4-bit or better with 8K context (32).
- At the same quant, speed is about the same (memory bandwidth 896 vs 960 GB/s).
- The RTX 5070 Ti launched at $749 vs $999 for the RTX 5080 (launch MSRP; street prices differ).
Specs that matter for LLMs
| RTX 5070 Ti | RTX 5080 | |
|---|---|---|
| Memory (VRAM) | 16 GB | 16 GB |
| Usable for a model (our estimate) | 15.2 GB | 15.2 GB |
| Memory bandwidth | 896 GB/s | 960 GB/s |
| Launch MSRP | $749 | $999 |
| Launch $ per GB of memory | $47 | $62 |
| Top 60 models that fit (4-bit or better, 8K ctx) | 32 | 32 |
| …of those at 8-bit or better | 30 | 30 |
| Largest top-60 model that fits | gpt-oss 20B | gpt-oss 20B |
| Offload headroom: more top-60 models via llama.cpp CPU offload (≥3 tok/s est., up to 128 GB RAM) | 18 more (1 more below 3 tok/s) | 18 more (1 more below 3 tok/s) |
Top 60 = the 60 most-downloaded open models we track. Launch MSRP from vendor announcements. How we calculate.
Estimated speed on popular models
Best quant that fits at 8K context, with a tokens/sec ceiling from memory bandwidth (real speed is typically 50–80% of it). "Offload" means part of the model sits in system RAM. Estimates, not benchmarks.
Models only one of them runs
None: of the top 60, both fit exactly the same models at 4-bit or better with 8K context.
More comparisons
RTX 4070 Ti Super vs RTX 5070 TiRTX 5070 Ti vs RX 9070 XTRTX 5080 vs RTX 5090RTX 4080 Super vs RTX 5080RTX 3090 vs RTX 5080RTX 5080 vs RX 7900 XTX All GPU comparisons
See every model on the RTX 5070 Ti or the RTX 5080, or try your own settings in the VRAM calculator.