llama.cpp on the AMD RX 6700 XT
On the AMD RX 6700 XT, llama.cpp runs Gemma 4 12B at 4-bit at a median 34.6 tokens per second for generation, from 1 plain run (8K context).
Speed by model on the AMD RX 6700 XT
Plain runs only: one device, one request, no speculative decoding, the whole model in memory. Each row is one model size at one bit class, with the other engines' plain runs of the same model on this card beside it. Context length and build still differ between runs, and each side shows its own.
| Model | llama.cpp | Other engines |
|---|---|---|
| Gemma 4 12B 4-bit | 34.6 t/sone report, 8K context | no other engine reported |
| Qwen3 8B 4-bit | 61.0 t/sone report, context not stated | no other engine reported |
4 llama.cpp reports on the AMD RX 6700 XT
Newest first. A run on several cards, with speculative decoding, with part of the model in system RAM or with batched requests says so, since each describes a different setup.
- Qwen3.6 35B · 3B active · Q4_K_XL30.0 t/s(speculative decoding, part in system RAM)
- Qwen3 8B · Q4_K_M61.0 t/s
- Gemma 4 12B · IQ4_NL34.6 t/s
- Qwen3.8 27B · UD-IQ3_XXS18.0 t/s(2 cards, speculative decoding)
Frequently asked
How fast is llama.cpp on the AMD RX 6700 XT?
On the AMD RX 6700 XT, llama.cpp runs Gemma 4 12B at 4-bit at a median 34.6 tokens per second for generation, from 1 plain run (8K context).