Model reference · open weights
LocalVQE is an open-weight audio or speech model from LocalAI-io, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.
About
LocalVQE [](https://huggingface.co/spaces/LocalAI-io/LocalVQE-demo) [](https://github.com/localai-org/LocalVQE) [](https://www.apache.org/licenses/LICENSE-2.0) Local Voice Quality Enhancement — compact neural models for acoustic echo cancellation (AEC), noise suppression (NS), and dereverberation of 16 kHz speech, running on commodity CPUs in real time. Causal and streaming (256-sample hop, 16 ms latency). - Try it: <https://huggingface.co/spaces/LocalAI-io/LocalVQE-demo - Source, build system, tests: <https://github.com/localai-org/LocalVQE This page hosts the published weights. Inference runs the GGML C++ engine on the GGUF files directly (build instructions on GitHub). Authors: Richard Palethorpe (richiejp) and Claude (Anthropic). LocalVQE is a streaming, CPU-tuned derivative of DeepVQE (Indenbom et al., Interspeech 2023). Models Speed is per 16 ms hop on a Ryzen 9 7900 (Zen4), 4 threads; RT = realtime factor (higher is faster than realtime). - Joint models (v1.2 / v1.3) clean echo, noise, and reverb in one pass. v1.3 is wider and filters noise better; v1.2 is ~1/4 the per-hop cost. - v1.4-AEC removes only the far-end echo and passes voice, room, and background through unchanged. It's a classical adaptive filter followed by a small neural mask. The 2.7K build is that filter alone — cheaper and gentler, but it can't remove heavily reverberant echo the way the mask can. - Every model needs a far-end reference signal (a loopback of what your speakers play) in addition to the mic. - bf16 GGUFs are ~12 % smaller with identical quality and speed; pick f32 unless download size matters. Compact line — GTCRN-AEC (for lower-power CPUs) A separate, much smaller second line of models for lower-power CPUs: a ~49 K-parameter GTCRN-AEC network — a distinct architecture based on GTCRN (Rong et al., ICASSP 2024) — paired with the project's DSP echo-cancellation front-end. The GGUFs are self-contained, so they run with the same single command as every other model. Two variants share the architecture: Whole-clip real-time factor on the real ggml graph, benchmarked on a Raspberry Pi 5 (one example of a low-power target; testgtcrn --bench, Cortex-A76, Ubuntu 24.04), parity-verifi
Summarised from the published model card. Read the full card on the HuggingFace links below.
Specifications
| Maker | LocalAI-io |
|---|---|
| Type | Audio & music |
| Variants | 1 |
| Runs with | pytorch |
| Released | 2026-04-17 |
| Popularity | 4k downloads / month |
| Likes | 66 |
| Licence | Open weights |
How it works
Variants
Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.
| Variant | Params | Precision | VRAM | Fits 16 GB | Weights |
|---|---|---|---|---|---|
| LocalVQE | — | BF16 | — | — | Weights ↗ |
Using it via the API
Once AxForge deploys localvqe for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (localvqe below is illustrative; you get the exact model name on deployment.)
$ curl -sS https://api.axforge.ai/v1/audio/transcriptions \ -H "Authorization: Bearer $AXFORGE_API_KEY" \ -F model="localvqe" -F file=@audio.mp3
Licence
Open weights under apache-2.0 — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗