Llama 3.1 8B Instruct › AMD Radeon 780M
Llama 3.1 8B Instruct on the AMD Radeon 780M
Best measured: 3.5 tokens/s generation and 110 tokens/s prompt processing (Q4_K_M, llamafile).
For comparison, our size-scaling estimate for this combination is 4.6 tok/s.
Every run
| Quant | Backend | Generate tok/s | Prompt tok/s | Test | Build / by / notes | Source |
|---|---|---|---|---|---|---|
| Q4_K_M | llamafile | 5.6 | 186 | LocalScore avg | LocalScore rank 182/422 | LocalScore (localscore.ai) |
| Q4_K_M | llamafile | 3.8 | 112 | LocalScore avg | LocalScore rank 226/422 | LocalScore (localscore.ai) |
| Q4_K_M | llamafile | 3.5 | 110 | LocalScore avg | LocalScore rank 236/422 | LocalScore (localscore.ai) |
| Q4_K_M | llamafile | 3.3 | 103 | LocalScore avg | LocalScore rank 241/422 | LocalScore (localscore.ai) |
| Q4_K_M | llamafile | 3.3 | 94 | LocalScore avg | LocalScore rank 251/422 | LocalScore (localscore.ai) |
tg128 = generating 128 tokens; pp512 = reading a 512-token prompt (llama-bench). LocalScore averages several prompt and output lengths, so its numbers run lower than tg128.
No card yet?
Try it on a rented GPU (RunPod)Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.
All results for the AMD Radeon 780M · Llama 3.1 8B Instruct on other hardware