# Moonshine Review

> Moonshine review: On-device streaming STT for live voice interfaces, from tiny edge models to Whisper Large v3-beating accuracy. Verified pricing, features.

![Moonshine logo](https://www.versusref.com/logos/moonshine.png)

On-device streaming STT for live voice interfaces, from tiny edge models to Whisper Large v3-beating accuracy

Among the 43 speech-to-text tools we track, Moonshine has the 38th-widest language coverage.

See pricing · Open source · facts verified Jul 20, 2026

## What we know about Moonshine

Moonshine sits in the speech-to-text apis category, where it is on-device streaming STT for live voice interfaces, from tiny edge models to Whisper Large v3-beating accuracy. We keep this Moonshine profile grounded in primary sources, each fact dated to when we last confirmed it.

Moonshine does not publish a public per-minute rate we have been able to verify, so the price figure here stays blank until we can confirm one. We would rather show nothing than a guessed number; if you have a current Moonshine quote, it is the fastest way for us to close that gap.

On capabilities, Moonshine covers word-level timestamps and self-host / on-prem option, and does not offer language auto-detection and websocket streaming api. Each of those is verified against Moonshine's own docs or dashboard, not marketing copy.

Among the 43 speech-to-text tools in our matrix, Moonshine leads with the 38th-widest language coverage; the fact sheet below has the raw numbers behind that placement.

In total we track 15 verified facts for Moonshine today, each linking the primary source it came from so you can check our work - and vendor claims we have not measured ourselves are labeled as such on the Moonshine fact sheet below.

## Fact sheet

### Capabilities

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| WER (vendor-claimed) | ~6.65 % WER | Jul 20 | [source](https://github.com/moonshine-ai/moonshine#when-should-you-choose-moonshine-over-whisper) |
| Streaming latency (vendor-claimed) | ~107 ms | Jul 20 | [source](https://github.com/moonshine-ai/moonshine#when-should-you-choose-moonshine-over-whisper) |
| Languages supported | 8 | Jul 20 | [source](https://github.com/moonshine-ai/moonshine#available-models) |
| Language auto-detection | ✗  No | Jul 20 | [source](https://github.com/moonshine-ai/moonshine#available-models) |
| Speaker diarization | ✓  Included | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |
| Word-level timestamps | ✓  Yes | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |

### Compliance & trust

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Self-host / on-prem option | ✓  Yes | Jul 20 | [source](https://moonshine.ai/) |
| Model weights license | MIT (English models); others non-commercial | Jul 20 | [source](https://github.com/moonshine-ai/moonshine#license) |

### Build experience

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Official SDKs | Python, JS (MoonshineJS), Swift, Java/Kotlin, C++ | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |
| Supported audio formats | WAV / raw PCM, mono, any sample rate | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |
| Websocket streaming API | ✗  No | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |

### Commercial

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Model size (parameters) | 26M-245M (Tiny to Medium Streaming) | Jul 20 | [source](https://github.com/moonshine-ai/moonshine#available-models) |
| Hardware to self-host | CPU-only; desktop, mobile, Raspberry Pi, MCUs, DSPs | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |
| Project maintenance status | Active - v0.0.69 released 2026-07-16 | Jul 20 | [source](https://github.com/moonshine-ai/moonshine/releases) |
| GitHub stars | 9,500 | Jul 20 | [source](https://github.com/moonshine-ai/moonshine) |

## Moonshine head-to-head

| Comparison | Record | Per use case |
| --- | --- | --- |
| [Moonshine vs OpenAI Whisper (API)](https://www.versusref.com/stt/moonshine-vs-whisper/) | won 3 · lost 2 · tied 2 | Call Centers: won; Developers: lost; Dictation: won; Medical: lost; Meetings: tie; Self-Hosted: won; Voice Agents: tie |

Source: https://www.versusref.com/stt/tools/moonshine/