ASMR-onnx is an open-weight model from Downflow, released under Apache License 2.0. Its published files total 4.5 GB.
Quantized graphs for the browser runtime: - llmint4/ - Qwen3.5-0.8B + ASMR LoRA, merged and exported, MatMulNBits int4 - ttsint4/ - F5-TTS DiT step + vocoder, int4 MatMul, convs in session dtype llmfp16/ and ttsfp16/ hold the fp16 exports they were quantized…
Model Card
By Downflow, published under apache-2.0, revision 5517194ab76b.
Quantized graphs for the browser runtime: - llmint4/ - Qwen3.5-0.8B + ASMR LoRA, merged and exported, MatMulNBits int4 - ttsint4/ - F5-TTS DiT step + vocoder, int4 MatMul, convs in session dtype llmfp16/ and ttsfp16/ hold the fp16 exports they were quantized from.
Read Downflow's full model card
WebASMR ONNX graphs
Quantized graphs for the browser runtime:
llm_int4/- Qwen3.5-0.8B + ASMR LoRA, merged and exported, MatMulNBits int4tts_int4/- F5-TTS DiT step + vocoder, int4 MatMul, convs in session dtype
llm_fp16/ and tts_fp16/ hold the fp16 exports they were quantized from.
Identity and Version
- Repository
- DownFlow/ASMR-onnx
- Publisher
- Downflow
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- 5517194ab76b22910df6841f0172211997c7da1f
- First published
- 2026-10-17
- Last updated
- 2026-10-18
Files and Weights
40 files, 4.5 GB in total. The weights are 16 files totalling 4.5 GB in onnx.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| base_fp16/dit_step.onnx | Weights | 685.8 MB | 6bb381bcff98 |
| base_fp16/vocoder.onnx | Weights | 27.2 MB | f05d15c17f4d |
| tts4_fp16/dit_step.onnx | Weights | 685.8 MB | 85d27d1e1a87 |
| tts4_fp16/vocoder.onnx | Weights | 27.2 MB | f05d15c17f4d |
| tts4_int4/dit_step.onnx | Weights | 407.1 MB | 76a263b934ab |
| tts4_int4/vocoder.onnx | Weights | 7.9 MB | 011f3a6866a0 |
| tts4v2_fp16/dit_step.onnx | Weights | 685.8 MB | 24ac112a687d |
| tts4v2_fp16/vocoder.onnx | Weights | 27.2 MB | f05d15c17f4d |
| tts4v2_int4/dit_step.onnx | Weights | 424.8 MB | 487211bcb7b4 |
| tts4v2_int4/vocoder.onnx | Weights | 9.2 MB | 53468ca1ddb1 |
| tts4v2_int4_b32/dit_step.onnx | Weights | 413.0 MB | 20d0e723fd30 |
| tts4v2_int4_b32/vocoder.onnx | Weights | 8.4 MB | ec8c557d2a33 |
| tts_fp16/dit_step.onnx | Weights | 685.8 MB | 6f4f03ae376b |
| tts_fp16/vocoder.onnx | Weights | 27.2 MB | f05d15c17f4d |
| tts_int4/dit_step.onnx | Weights | 407.1 MB | e0159ef4b5e5 |
| tts_int4/vocoder.onnx | Weights | 7.9 MB | 011f3a6866a0 |
| base_fp16/manifest.json | Configuration | 1.8 KB | — |
| sweep/sweep.json | Configuration | 1.6 KB | — |
| tts4_fp16/manifest.json | Configuration | 1.9 KB | — |
| tts4_int4/manifest.json | Configuration | 1.9 KB | — |
| tts4_int4/quantization.json | Configuration | 826 B | — |
| tts4v2_fp16/manifest.json | Configuration | 1.8 KB | — |
| tts4v2_int4/manifest.json | Configuration | 1.8 KB | — |
| tts4v2_int4/quantization.json | Configuration | 403 B | — |
| tts4v2_int4_b32/manifest.json | Configuration | 1.8 KB | — |
| tts4v2_int4_b32/quantization.json | Configuration | 403 B | — |
| tts_fp16/manifest.json | Configuration | 2.0 KB | — |
| tts_int4/manifest.json | Configuration | 2.0 KB | — |
| tts_int4/quantization.json | Configuration | 830 B | — |
| README.md | Documentation | 360 B | — |
| sweep/sweep.csv | Other | 286 B | — |
| .gitattributes | Repository | 1.5 KB | — |
| base_fp16/vocab.txt | Tokenizer | 11.3 KB | — |
| tts4_fp16/vocab.txt | Tokenizer | 11.3 KB | — |
| tts4_int4/vocab.txt | Tokenizer | 11.3 KB | — |
| tts4v2_fp16/vocab.txt | Tokenizer | 11.3 KB | — |
| tts4v2_int4/vocab.txt | Tokenizer | 11.3 KB | — |
| tts4v2_int4_b32/vocab.txt | Tokenizer | 11.3 KB | — |
| tts_fp16/vocab.txt | Tokenizer | 11.3 KB | — |
| tts_int4/vocab.txt | Tokenizer | 11.3 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 4.5 GB
Released by Downflow through its official repository on Hugging Face. Read the license.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 4.5 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About ASMR-onnx
Can I use ASMR-onnx commercially?
Yes. ASMR-onnx is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.