The open-source alternatives to Deepgram are Whisper and its ports (whisper.cpp, WhisperX), NVIDIA's Parakeet through NeMo Speech, Moonshine, Vosk, sherpa-onnx and Speaches. For a recording, Whisper's family is the default: MIT weights, six model sizes, and WhisperX adds word timestamps and speaker labels. For live audio, Moonshine and Vosk are built for streaming on small hardware, and sherpa-onnx runs both modes on a CPU or a phone. Speaches wraps these in an OpenAI-compatible server. A team switches when audio cannot leave its network or when per-minute pricing outgrows a GPU.
Deepgram, in the vendor's own words: "voice AI built for real conversations", with a speech-to-text model it describes as conversation-aware, built to understand turn-taking and handle interruptions, by its own description. deepgram.com.
| Project | Openness | Licence | Runs as | Platforms | Self-host | Hosted option | Last release | Stars |
|---|---|---|---|---|---|---|---|---|
| Whisper | OSI | MIT | pip, CLI, Python library | Linux, macOS, Windows (ffmpeg required) | Yes | No | Jun 26, 2025 | 109,222 |
| whisper.cpp | OSI | MIT | C/C++ library, CLI, Server (OpenAI-like API), Docker | macOS (Intel and Apple Silicon), Linux, FreeBSD, Windows, iOS, Android, WebAssembly, Raspberry Pi | Yes | No | Sep 11, 2026 | 53,711 |
| WhisperX | OSI | BSD-2-Clause | pip, CLI, Python library | Linux and Windows (CUDA 12.8), macOS and CPU (int8 mode) | Yes | No | May 25, 2026 | 24,069 |
| Parakeet (NeMo Speech) | Open weights | Apache-2.0 | pip, Docker, Python library | Linux with an NVIDIA GPU (recommended); a CPU path through Transformers is documented on the model card | Yes | No | Aug 7, 2026 | 18,465 |
| Vosk | OSI | Apache-2.0 | Language bindings (Python, Java, Node.js, C#, C++, Rust, Go) | Raspberry Pi, Android, servers; desktop OS list not stated | Yes | No | Apr 22, 2024 | 15,131 |
| sherpa-onnx | OSI | Apache-2.0 | Libraries in 12 languages, prebuilt Android and Flutter apps, WebAssembly, websocket server | Linux, macOS, Windows, Android, iOS, HarmonyOS, Node.js, WebAssembly, Raspberry Pi and named boards | Yes | No | Sep 10, 2026 | 14,808 |
| Moonshine | OSI | See LICENSE | pip, CLI, C library, JavaScript/WASM, mobile SDKs | Python, JavaScript/WASM, iOS, Android, macOS, Linux, Windows, Raspberry Pi | Yes | No | Aug 24, 2026 | 11,089 |
| Speaches | OSI | MIT | Docker, Server (OpenAI-compatible API) | Docker on GPU or CPU | Yes | No | Dec 27, 2025 | 3,665 |
Licence, release and star figures read from GitHub on September 16, 2026. Openness labels: The code, and the weights where the project is a model, carry a licence on the OSI approved list. The code is public, and the licence limits what you may do with it. Read the LICENSE file before commercial use. The model weights can be downloaded and run. The training data or code is closed, or the weights carry a restriction the code does not.
OpenAI's general-purpose speech recognition model, released with code and weights under MIT, in six sizes from 39 million to 1.55 billion parameters, with multilingual transcription, translation to English and language detection.
pip, CLI, Python library · Linux, macOS, Windows (ffmpeg required) · Runs on your own machines
A dependency-free C and C++ port of Whisper under MIT that runs on CPUs, Apple Silicon and phones, with quantised models, an HTTP server and bindings for a dozen languages.
C/C++ library, CLI, Server (OpenAI-like API), Docker · macOS (Intel and Apple Silicon), Linux, FreeBSD, Windows, iOS, Android, WebAssembly, Raspberry Pi · Runs on your own machines
Batched Whisper transcription under BSD-2-Clause with word-level timestamps from forced alignment and speaker labels from pyannote, the closest open-source match to a hosted transcription API's output.
pip, CLI, Python library · Linux and Windows (CUDA 12.8), macOS and CPU (int8 mode) · Runs on your own machines · Weights: Ships no weights of its own; the speaker diarization model from pyannote is CC-BY-4.0 and gated behind a Hugging Face agreement
NVIDIA's speech models and toolkit: Apache-2.0 code, and the Parakeet TDT 0.6B v3 model under CC-BY-4.0 with 25 European languages, automatic language detection, punctuation and word timestamps.
pip, Docker, Python library · Linux with an NVIDIA GPU (recommended); a CPU path through Transformers is documented on the model card · Runs on your own machines · Weights: CC-BY-4.0 on the Parakeet weights; the NeMo Speech code is Apache-2.0
An offline speech recognition toolkit under Apache-2.0 with streaming, models of about 50 MB, 20+ languages, a reconfigurable vocabulary and speaker identification, from small boards to clusters.
Language bindings (Python, Java, Node.js, C#, C++, Rust, Go) · Raspberry Pi, Android, servers; desktop OS list not stated · Runs on your own machines
A runtime under Apache-2.0 from the next-generation Kaldi team that runs speech to text, text to speech, speaker diarization and voice activity detection locally on CPUs, NPUs and phones, with no internet connection.
Libraries in 12 languages, prebuilt Android and Flutter apps, WebAssembly, websocket server · Linux, macOS, Windows, Android, iOS, HarmonyOS, Node.js, WebAssembly, Raspberry Pi and named boards · Runs on your own machines
An on-device speech toolkit under MIT built for live streaming, with speech-to-text models trained from scratch in sizes down to 1 MB, one library across Python, the browser, phones and desktops.
pip, CLI, C library, JavaScript/WASM, mobile SDKs · Python, JavaScript/WASM, iOS, Android, macOS, Linux, Windows, Raspberry Pi · Runs on your own machines · Weights: MIT for every streaming model and every English model; thirteen legacy non-English non-streaming models stay under a non-commercial community licence
An OpenAI-compatible server under MIT for streaming transcription, translation and text to speech, loading Whisper, Kokoro and Piper models on demand, described by its maintainers as Ollama for speech models.
Docker, Server (OpenAI-compatible API) · Docker on GPU or CPU · Runs on your own machines
For recorded audio, OpenAI's Whisper (MIT code and weights), or WhisperX if you need word timestamps and speaker labels, or whisper.cpp if you need it on a CPU, a Mac or a phone. For live streaming audio, Moonshine (MIT, built for low latency on device) or Vosk (Apache-2.0, streaming API, models of about 50 MB). For an OpenAI-compatible server around these, Speaches.
Moonshine is designed for live streaming and says it works while the user is still talking. Vosk has a streaming API with what its README calls zero-latency response. sherpa-onnx supports both streaming and non-streaming models. whisper.cpp has a stream example that transcribes continuously from the microphone. NVIDIA's Nemotron ASR Streaming model runs through NeMo Speech. Whisper itself processes a sliding 30-second window.
WhisperX does both: word timestamps by alignment with a wav2vec2 model, and speaker labels through pyannote, whose model is CC-BY-4.0 and needs a Hugging Face token. whisper.cpp has experimental word timestamps and experimental speaker turn markers. NVIDIA's Parakeet TDT reports word and segment timestamps. Vosk offers speaker identification.
Yes. whisper.cpp is built for it and runs fully on the GPU on Apple Silicon through Metal. Vosk runs on a Raspberry Pi. Moonshine ships models down to 1 MB for small devices. sherpa-onnx runs on x86, ARM and RISC-V CPUs. WhisperX documents an int8 CPU mode. NeMo Speech recommends an NVIDIA GPU for inference, and the Parakeet card documents a CPU path through Transformers.
Not a Deepgram-shaped one on this page. Speaches and whisper.cpp's server both expose an OpenAI-compatible API, so code written against OpenAI's transcription endpoint works with a base URL change. Code written against Deepgram's own SDK needs its calls rewritten.
Whisper's code and weights are MIT. whisper.cpp is MIT. WhisperX is BSD-2-Clause, but its diarization model is CC-BY-4.0 and gated. Moonshine's code, every streaming model and every English model are MIT; thirteen legacy non-English models are under a non-commercial community licence. Parakeet weights are CC-BY-4.0, which permits commercial use with attribution. Vosk and sherpa-onnx code are Apache-2.0; their model licences are not stated in the READMEs.
This page does not compare accuracy, because no shared measurement exists here. The projects publish their own figures: Moonshine claims accuracy from above Whisper Large V3 down to 1 MB models, and NVIDIA's Canary-Qwen reports a 5.63 percent word error rate on the English Open ASR Leaderboard. Both are the projects' own claims.
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.