Model reference · open weights
MOSS-Transcribe is an open-weight audio or speech model from OpenMOSS-Team, 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
MOSS-Transcribe-preview-2B MOSS-Transcribe-preview-2B is an English speech-to-text model that pairs a Qwen3-1.7B-base language-model backbone with a Qwen3-Omni-MoE audio encoder. A gated-MLP adapter projects audio features into the language-model embedding space. The model is trained on public English ASR corpora and fine-tuned with reinforcement learning on the Open ASR Leaderboard training splits. The model has approximately 2.4B parameters and is distributed as a single bfloat16 safetensors shard of approximately 4.84 GB. Model Details - Developed by: OpenMOSS Team - Model type: Automatic Speech Recognition / speech-to-text model - Language: English - License: Apache-2.0 - Library: Transformers - Backbone: Qwen3-1.7B-base, 28 layers, hidden size 2048 - Audio encoder: Qwen3-Omni-MoE audio encoder - Adapter: Gated-MLP adapter, hidden size 8192 - Parameter size: approximately 2.4B - Checkpoint format: bfloat16 safetensors Intended Use This model is intended for English automatic speech recognition, including transcription of English speech audio for research and evaluation purposes. Evaluation Evaluated on the Open ASR Leaderboard test sets. Predictions are produced with greedy decoding (numbeams=1, maxnewtokens=512), a single dataset-agnostic chat template, and scored with the leaderboard's standardized scoring (English normalizer + word-level edit distance with compound merging). TED-LIUM is not currently part of the leaderboard run and is therefore excluded. Inference Audio Frontend - Sample rate: 16 kHz - Features: Whisper log-mel filterbank - Mel bins: 128 - FFT size: 400 - Hop length: 160 Training The model was trained on public English ASR corpora and fine-tuned with reinforcement learning on the Open ASR Leaderboard training splits. Limitations The model is designed for English ASR. It may perform worse on non-English speech, heavy accents, noisy recordings, overlapping speakers, far-field audio, domain-specific terminology, or audio conditions that differ significantly from the training and evaluation data. The output should be manually reviewed before use in high-stakes settings. Citation License This model is released under th
Summarised from the published model card. Read the full card on the HuggingFace links below.
Specifications
| Maker | OpenMOSS-Team |
|---|---|
| Type | Audio & music |
| Parameters (lead) | 2.4B |
| Variants | 1 |
| Runs with | transformers |
| Released | 2026-06-26 |
| Popularity | 1k downloads / month |
| Likes | 50 |
| 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 |
|---|---|---|---|---|---|
| MOSS-Transcribe-preview-2B | 2.4B | BF16 | ~5.6 GB | ✓ | Weights ↗ |
Benchmarks
As published on the model card — the maker's own numbers, not measured by AxForge.
| Task | Dataset | Metric | Score |
|---|---|---|---|
| automatic-speech-recognition | Open ASR Leaderboard | Average WER | 4.87 |
Using it via the API
Once AxForge deploys moss-transcribe for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (moss-transcribe 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="moss-transcribe" -F file=@audio.mp3
Details
Languages
Trained / evaluated on
Tags
Licence
Open weights under apache-2.0 — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗