RTX 5070 Ti for local LLMs
6 measured results from 3 builds, each linked to its source.
| Memory | 16 GB per card |
|---|---|
| Memory bandwidth | 896 GB/s (memory bandwidth as listed on llamabench) |
| Typical EU price | ≈€1,188 new (PNY GeForce RTX 5070 Ti 16 GB GDDR7, billiger.de, 2026-10-16) |
Can it run? Gemma 4 12B on RTX 5070 Ti Qwen3.8 27B on 2x RTX 5070 Ti Qwen3.8 27B on RTX 5070 Ti Qwen3.8 Flash-Next on 2x RTX 5070 Ti gpt-oss 20B on RTX 5070 Ti
All builds with the RTX 5070 Ti
| Build | GPU(s) | Memory GB (total) | Results |
|---|---|---|---|
| Dual RTX 5070 Ti on Ryzen 9 9900X (capy-flash maintainer rig) | 2x RTX 5070 Ti 16 GB | 32 | 4 |
| Single RTX 5070 Ti 16GB | RTX 5070 Ti | 16 | 1 |
| RTX 5070 Ti: 2026 models (llamabench community) | RTX 5070 Ti | 16 | 1 |
All measured results
(MTP, DFlash, DSpark and similar draft-and-verify methods give much higher tok/s on predictable text and are not comparable 1:1 with plain decode)
| Model | Build | Ran on | Quant | Engine | tok/s | Details |
|---|---|---|---|---|---|---|
| Gemma 4 12B | RTX 5070 Ti: 2026 models (llamabench community) | RTX 5070 Ti | llama-bench tg128 (quant per result) | llama.cpp | 87.6 | |
| Qwen3.8 27B | Dual RTX 5070 Ti on Ryzen 9 9900X (capy-flash maintainer rig) | 2x RTX 5070 Ti | 3-bit GSQ | capy-flash, 2 GPUs, MTP, decode at 8k prompt | 230 | ⚡ spec-decode (MTP)context: 8k prompt📏 Maintainer/user decode tok/s from the capy-flash README |
| Qwen3.8 27B | Dual RTX 5070 Ti on Ryzen 9 9900X (capy-flash maintainer rig) | RTX 5070 Ti | 3-bit GSQ | capy-flash (1 GPU), MTP, decode at 8k prompt | 139 | ⚡ spec-decode (MTP)context: 8k prompt📏 Maintainer/user decode tok/s from the capy-flash README |
| Qwen3.8 Flash-Next | Dual RTX 5070 Ti on Ryzen 9 9900X (capy-flash maintainer rig) | 2x RTX 5070 Ti | GSQ-RCO (64 GB edition) | capy-flash v0.7.0, 2 GPUs, agentic decode | 182 | 📏 Maintainer/user decode tok/s from the capy-flash READMEpartly in system RAM |
| Qwen3.8 Flash-Next | Dual RTX 5070 Ti on Ryzen 9 9900X (capy-flash maintainer rig) | 2x RTX 5070 Ti | NVFP4 | capy-flash v0.6.0, 2 GPUs, agentic decode | 155 | 📏 Maintainer/user decode tok/s from the capy-flash READMEpartly in system RAM |
| gpt-oss 20B | Single RTX 5070 Ti 16GB | RTX 5070 Ti | MXFP4 | llama.cpp | 189.45 |
All RTX 5070 Ti builds: see also all GPUs · best LLMs by memory size · download the dataset.
FAQ
How fast is the RTX 5070 Ti for local LLMs?
The fastest plain-decode result we have is 189.45 tok/s on gpt-oss 20B (MXFP4, llama.cpp) in Single RTX 5070 Ti 16GB. Speed depends heavily on the model, quant and engine, so compare like with like in the table. RigBase lists 6 measured results from 3 builds.
How much memory does the RTX 5070 Ti have?
16 GB per card. Memory bandwidth: 896 GB/s (memory bandwidth as listed on llamabench).
Which models have been measured on the RTX 5070 Ti?
Gemma 4 12B, Qwen3.8 27B, Qwen3.8 Flash-Next, gpt-oss 20B.
Where do these numbers come from?
Every row links to a build page with its original public source (benchmark logs, READMEs, forum posts). RigBase does not re-run the benchmarks. Results marked ⚡ use speculative decoding and are not comparable 1:1 with plain decode.
Referral link: if you sign up with Google and add $10, Runpod gives you and RigBase bonus credit (terms). It never affects our data.