--- title: OctoASR-1.7B-Instruct-1.0-MLX-8bit canonical_url: "https://www.modelscope.ai/models/Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit" md_url: "https://www.modelscope.ai/models/Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit.md" repository: Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit last_updated: 2026-07-27 pipeline_tag: auto-speech-recognition tasks: - auto-speech-recognition model_type: - qwen3_asr architectures: - Qwen3ASRForConditionalGeneration parameters: 801.5M tensor_type: - U32 - BF16 library_name: - safetensors - pytorch frameworks: - Pytorch downloads: 19 stars: 0 --- # OctoASR-1.7B-Instruct-1.0-MLX-8bit > OctoASR-1.7B-Instruct-1.0-MLX-8bit - An open-source model by Mininglamp2718 on ModelScope. OctoASR is a local speech recognition service for vertical domains, deeply optimized for Apple Silicon via Cider. Purpose-built for Internet or IT office scenarios, it… - **Repository**: Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit - **Tasks**: auto-speech-recognition - **Parameters**: 801.5M - **Downloads**: 19 - **Stars**: 0 - **Last updated**: 2026-07-27 Source: https://www.modelscope.ai/models/Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit ---

OctoASR

MLX License Stars HuggingFace ModelScope CN ModelScope AI

δΈ­ζ–‡ | English

--- ## Introduction **OctoASR** is a local speech recognition service for vertical domains, deeply optimized for Apple Silicon via [Cider](https://github.com/Mininglamp-AI/cider). Purpose-built for **Internet or IT office** scenarios, it is closely adapted to high-frequency workplace use cases such as meeting notes, technical discussions, product reviews, and engineering dictation. Through targeted optimization on domain data, OctoASR accurately recognizes English terms, acronyms, and product names (e.g. `Kubernetes`, `FastAPI`, `PRD`, `Code Review`) as well as mixed Chinese-English speech, effectively addressing the term-misrecognition and code-switching segmentation issues common to general-purpose models β€” so transcripts come out clear, domain-aware, and accurate. The service runs fully locally, works out of the box, and keeps audio and transcript data on your machine. Core capabilities: - 🎯 **Vertical-domain optimization** β€” optimized on Internet or IT office data; accurate on English terms, acronyms, product names, and mixed Chinese-English speech. - 🍎 **Native Apple Silicon** β€” MLX-based local inference on M-series chips, further optimized with our in-house acceleration framework Cider. - πŸ”’ **Fully local, privacy-first** β€” audio and transcripts never leave your machine. - βœ‚οΈ **VAD segmentation** β€” optional FSMN VAD splits long audio and transcribes segment by segment. - 🧩 **Pluggable engines** β€” supports Fun-ASR-Nano, Qwen3-ASR and more base models, switchable with one command. - 🏷️ **@Mention replacement** β€” auto-fix nicknames and transliterated names in transcripts via a visual page. See [Mentions](docs/mentions/README.md). - ⚑ **One-command start** β€” install via `brew install`, then `octoasr start`. ---

Changelog Β· Models Β· Installation Β· Usage Β· API Β· License Β· Acknowledgments Β·

--- ## Changelog See the full release history on the **[Releases](https://github.com/Mininglamp-AI/octoasr/releases)** page. - **2026-06-15** β€” Release the first ASR model built on Qwen3-ASR and tailored for internet office scenarios, supporting handwritten-style transcription output and accurate recognition of industry terminology. - **2026-06-09** β€” Added @mention replacement with a visual management page (`octoasr mentions`) for editing nickname β†’ canonical-name mappings; spoken "θ‰Ύη‰Ή" is normalized to `@` before replacement. (v0.1.15 fixes packaging so the web page ships in the Homebrew build.) - **2026-05-29** β€” Released the first ASR model for internet office scenarios, with written-style transcription output and accurate recognition of industry-specific terminology. - **2026-05-26** β€” First release: FastAPI transcription service, FunASR-Nano engine, FSMN VAD, hotword extraction, session logging. --- ## Models OctoASR uses a pluggable engine design and supports several mainstream ASR base models. Switch with a single command: `octoasr model use `. | Model | Base model | Quant | Size | Languages | Links | | --- | --- | --- | --- | --- |-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | **OctoASR-1.7B** (default) | Qwen3-ASR | 8bit | 1.7 GB | ZH / EN | [πŸ€—](https://huggingface.co/Mininglamp-2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit) Β· [πŸ€–](https://www.modelscope.cn/models/Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit) Β· [🌟](https://www.modelscope.ai/models/Mininglamp2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit) | | **OctoASR-0.8B** | Fun-ASR-Nano | 8bit | 0.8 GB | ZH / EN | [πŸ€—](https://huggingface.co/Mininglamp-2718/OctoASR-0.8B-Instruct-1.0-MLX-8bit) Β· [πŸ€–](https://www.modelscope.cn/models/Mininglamp2718/OctoASR-0.8B-Instruct-1.0-MLX-8bit) Β· [🌟](https://www.modelscope.ai/models/Mininglamp2718/OctoASR-0.8B-Instruct-1.0-MLX-8bit) | > The model is downloaded automatically from HuggingFace or ModelScope (China mirror); the source is chosen by network environment on first run. --- ## Installation ### Option 1: Homebrew (recommended) ```bash brew tap Mininglamp-AI/tap brew install octoasr # Start (first run auto-initializes + downloads the default model) octoasr start octoasr doctor # environment check ``` ### Option 2: From source ```bash # 1. Dependency: ffmpeg (decodes non-WAV audio) brew install ffmpeg # 2. Clone + install git clone https://github.com/Mininglamp-AI/OctoASR.git cd octoasr python3 -m venv .venv && source .venv/bin/activate pip install -U pip pip install -e . # 3. Download the model hf download Mininglamp-2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit \ --local-dir models/Mininglamp-2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit hf download Mininglamp-2718/fsmn-vad-mlx \ --local-dir models/Mininglamp-2718/fsmn-vad-mlx # Behind a China mirror: # HF_ENDPOINT=https://hf-mirror.com hf download ... # 4. Start the server python3 server.py \ --model-path models/Mininglamp-2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit \ --vad-model-path models/Mininglamp-2718/fsmn-vad-mlx \ --host 0.0.0.0 --port 8787 --load-on-startup ``` **Requirements:** macOS (Apple Silicon) Β· Python 3.10+ Β· `ffmpeg` / `ffprobe` on `PATH`. --- ## Usage The example below shows **audio translation**: transcribe speech and translate it into Chinese. ### CLI (recommended) ```bash # On first run, the service auto-initializes and downloads the default model octoasr start # Transcribe / translate an audio file octoasr transcribe assets/BAC009S0764W0129.wav ``` ### Python API ```python from core.auto_model import AutoModel model = AutoModel( model="models/Mininglamp-2718/OctoASR-1.7B-Instruct-1.0-MLX-8bit", vad_model="models/Mininglamp-2718/fsmn-vad-mlx", # optional: auto-segment long audio ) text = model.generate( "assets/BAC009S0764W0129.wav", task="translate", # translation task target_language="zh", # target language: Chinese merge_vad=True, ) print(text) # -> "η”šθ‡³ε‡ΊηŽ°δΊ€ζ˜“ε‡ δΉŽεœζ»žηš„ζƒ…ε†΅" ``` ### HTTP API ```bash curl -X POST http://127.0.0.1:8787/v1/voice/transcribe \ -F "audio=@assets/BAC009S0764W0129.wav" \ -F "personal_context=## Terms\n- FastAPI\n- Kubernetes" \ -F "mode=smart" ``` ```json { "status": 200, "text": "transcribed text", "m": "octoasr", "engine": "mlx" } ``` > Full API fields, limits and auth are documented under [API](#en-api). ### @Mention replacement Auto-replace casual nicknames and transliterated names in transcripts with the canonical spelling you want (e.g. `@小明` β†’ `@Xiaoming`). Manage entries on a visual web page β€” no JSON editing required: ```bash octoasr start # Start the service (if not running) octoasr mentions # Open the management page in your browser ``` > πŸ“– Full guide: [Mentions](docs/mentions/README.md) --- ## API ### `POST /v1/voice/transcribe` Transcribe a single uploaded audio file. Request type: `multipart/form-data`. | Field | Type | Required | Description | | --- | --- | --- | --- | | `audio` | file | yes | Audio file. Supported: `.wav` `.mp3` `.ogg` `.webm` `.m4a` `.flac` | | `context_text` | string | no | Existing text for append/edit modes; last 5000 chars kept | | `chat_context` | string | no | Chat context; last 20000 chars kept | | `personal_context` | string | no | Personal correction / hotword context; last 10000 chars kept | | `member_context` | string | no | Member context; last 5000 chars kept | | `mode` | string | no | `smart` / `append_only` / `edit_only`, default `smart` | **Limits:** default max file `30 MiB`, max duration `660` s; `edit_only` requires `context_text`. ### `GET /v1/voice/config` Returns current service limits and engine metadata. ```bash curl http://127.0.0.1:8787/v1/voice/config ``` ### Authentication Disabled by default. If started with `--auth-token`, requests must carry `Authorization: Bearer `. ```bash python3 server.py --model-path --auth-token "$OCTOASR_TOKEN" ``` --- ## License Released under the [MIT License](LICENSE). Copyright (c) 2026 MININGLAMP Technology. --- ## Acknowledgments OctoASR would not be possible without these excellent open-source projects: - [**MLX**](https://github.com/ml-explore/mlx) & [**mlx-audio**](https://github.com/Blaizzy/mlx-audio) β€” Apple's machine-learning framework and audio toolkit, the foundation of OctoASR's local inference. - [**FunASR / FunAudioLLM**](https://github.com/modelscope/FunASR) β€” source of Fun-ASR-Nano and FSMN-VAD, providing strong Chinese speech recognition. - [**Qwen3**](https://github.com/QwenLM/Qwen3) β€” the base model behind the Qwen3-ASR engine. - [**mlx-community**](https://huggingface.co/mlx-community) β€” high-quality MLX quantized models. - [**ModelScope**](https://github.com/modelscope/modelscope) & [**Hugging Face**](https://huggingface.co/) β€” model hosting and distribution. - [**FastAPI**](https://github.com/fastapi/fastapi) β€” high-performance web framework. Thanks to everyone contributing to the open-source speech recognition community.