selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

Edge0 logo

Can the Radeon RX 6700 XT run Audio8-TTS-Realistic-0.6b?

Audio8-TTS-Realistic-0.6b·601M·Radeon RX 6700 XT·12 GB VRAM

Saved in this browser, used on every model page.
Yes

Yes. The Radeon RX 6700 XT runs Audio8-TTS-Realistic-0.6b fully in VRAM at FP16 / BF16 (about 2.6 GB of 12 GB).

best quant
FP16 / BF16
needs
2.6 GB
usable
12 GB

Estimates from the model's size.Try other contexts on the Audio8-TTS-Realistic-0.6b page →

Every quantisation of Audio8-TTS-Realistic-0.6b on the Radeon RX 6700 XT

QuantMemoryFit
Q4_K_M 0.9 GBfits
Q8_0 2.1 GBfits
FP16 / BF16 pick2.6 GBfits

Speed is a bandwidth-based estimate for generating text (384 GB/s); prompt processing and long contexts are slower.

Run Audio8-TTS-Realistic-0.6b on the Radeon RX 6700 XT

Transformers · Run with the text-to-speech pipelinedocs
from transformers import pipeline

pipe = pipeline("text-to-speech", model="Edge0/Audio8-TTS-Realistic-0.6b")
result = pipe("Self-hosting puts you in control of your data.")
print(result)

TTS models that fit the Radeon RX 6700 XT comfortably

All that fit →
ModelParamsBest quantMemoryLicence
k2-fsa logoOmniVoicek2-fsa
613MFP16 / BF16≈1.9 GBFits in memoryCheck licence
601MFP16 / BF16≈1.8 GBFits in memoryCommercial OK
ResembleAI logochatterboxResembleAI
500MFP16 / BF16≈1.6 GBFits in memoryCommercial OK
500MFP16 / BF16≈1.6 GBFits in memoryCommercial OK
601MFP16 / BF16≈1.8 GBFits in memoryCommercial OK
915MFP16 / BF16≈2.5 GBFits in memoryCommercial OK

Other GPUs that run Audio8-TTS-Realistic-0.6b

Frequently asked questions

Can the Radeon RX 6700 XT run Audio8-TTS-Realistic-0.6b?
Yes. The Radeon RX 6700 XT runs Audio8-TTS-Realistic-0.6b fully in VRAM at FP16 / BF16 (about 2.6 GB of 12 GB).
Which quantisation of Audio8-TTS-Realistic-0.6b should I use on the Radeon RX 6700 XT?
FP16 / BF16: 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.