Can the RTX A6000 run Apriel-Voice-27B-preview?
Apriel-Voice-27B-preview·27.2B·RTX A6000·48 GB VRAM
Yes. The RTX A6000 runs Apriel-Voice-27B-preview fully in VRAM at Q8_0 (about 31 GB of 48 GB), at roughly 15 tokens/s.
- best quant
- Q8_0
- needs
- 31 GB
- usable
- 48 GB
- est. speed
- ~15 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the Apriel-Voice-27B-preview page →
Every quantisation of Apriel-Voice-27B-preview on the RTX A6000
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 11 GB | fits | ~47 tok/s |
| Q3_K_M | 15 GB | fits | ~32 tok/s |
| Q4_K_M | 19 GB | fits | ~26 tok/s |
| Q5_K_M | 21 GB | fits | ~22 tok/s |
| Q6_K | 24 GB | fits | ~19 tok/s |
| Q8_0 pick | 31 GB | fits | ~15 tok/s |
| FP16 / BF16 | 56 GB | offload | ~3 tok/s |
Speed is a bandwidth-based estimate for generating text (768 GB/s); prompt processing and long contexts are slower.
Measured speed on the RTX A6000
No community reports for this pair yet.
Report your own speed →Omni models that fit the RTX A6000 comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 35.3BA3B | Q8_0 | ≈33 GBFits in memory | ~133 tok/s | |
| 24.3B | Q8_0 | ≈26 GBFits in memory | ~16 tok/s | |
| 31.7BA3B | Q8_0 | ≈33 GBFits in memory | ~133 tok/s | |
| 33BA3B | Q8_0 | ≈34 GBFits in memory | ~133 tok/s | |
| 31.7BA3B | Q8_0 | ≈34 GBFits in memory | ~133 tok/s | |
| 17.5B | Q8_0 | ≈20 GBFits in memory | ~23 tok/s |
Other GPUs that run Apriel-Voice-27B-preview
Frequently asked questions
- Can the RTX A6000 run Apriel-Voice-27B-preview?
- Yes. The RTX A6000 runs Apriel-Voice-27B-preview fully in VRAM at Q8_0 (about 31 GB of 48 GB), at roughly 15 tokens/s.
- How fast does Apriel-Voice-27B-preview run on the RTX A6000?
- An estimate from memory bandwidth gives about 15 tokens/s at Q8_0.
- Which quantisation of Apriel-Voice-27B-preview should I use on the RTX A6000?
- Q8_0: the largest one that fits. Expected quality: lossless compared with the full-precision weights.
Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.