What is AudioShake?
AudioShake is an AI audio separation platform that breaks a finished recording into its component parts, known as stems. It can pull vocals, drums, bass, guitar, piano and other instruments out of a song, separate dialogue, music and effects in film and TV audio, and isolate individual speakers in a conversation. It also offers lyric transcription with word-level alignment and music detection for copyright checks.
You upload a file to AudioShake Studio in the browser, or send audio to the REST API, choose the target stems and download the results. A developer SDK adds real-time, low-latency separation. AudioShake was co-founded by CEO Jessica Powell and CTO Luke Miner, and Axios reports a founding year of 2020 and a $14 million Series A in October 2025, bringing total funding to about $19 million. The platform was live in October 2026.
Key Features
- Music stem separation: AudioShake splits songs into more than a dozen instrument targets, including lead vocals, backing vocals, drums, bass, electric and acoustic guitar, piano, keys, strings and wind. A drum kit model can further divide drums into six separate stems.
- Film and TV post-production: AudioShake separates dialogue, music and effects from mixed soundtracks for dubbing and editing. The music and effects target removes dialogue, which helps studios prepare localised versions of a film or show.
- Multi-speaker separation: AudioShake outputs one stem per speaker from overlapping conversation audio. A cheaper diarisation model returns speaker identities and timings, which suits transcription and captioning workflows.
- Speech cleanup: AudioShake removes background noise, hum and room reverb from speech with separate denoise and dereverb models. Studio groups these under Speech Recovery for podcasts, interviews and field recordings.
- Lyric transcription and alignment: AudioShake generates line-level lyric transcripts and word-level timestamps from songs. This supports karaoke features, synced lyrics and caption files, which you can export as JSON or TXT.
- Copyright compliance tools: AudioShake detects where music appears in a file, identifies tracks and can remove background music while keeping speech and effects. Broadcasters and creators use this to clear unlicensed music from videos.
- REST API with credit billing: AudioShake bills API jobs per minute of source audio for each target model, rounded up to the nearest minute. Most instrument stems cost 1.0 credit per minute, while multi-voice separation costs 10.0.
- Real-time SDK: AudioShake offers an SDK for real-time, low-latency separation, and its FAQ says many models can run on-device. Developers can build stem features into apps, games or hardware rather than sending everything to the cloud.
- High-resolution output: AudioShake supports audio up to 192kHz and exports WAV, MP3, AAC, FLAC, AIFF and PCM. Studio accepts uploads up to 2GB, including MP4 and MOV video files.
Pros
- Proven with major rights holders: AudioShake reports work with Universal Music Group, Warner Music Group, Disney Music Group and BMG. Digital Music News counted more than 40 enterprise clients at its Series A, which signals trust from demanding customers.
- Award-winning separation: AudioShake won Sony’s Demixing Challenge against about 40 competitors, according to its FAQ. Its team page also cites the AWS re:Invent Startup Competition and a TIME Best Inventions listing.
- Covers music, film and speech: AudioShake handles songs, soundtracks and conversations in one platform rather than focusing on music alone. That breadth suits post-production teams who move between dubbing, cleanup and remix work.
- Free way to start: AudioShake Studio offers a free plan for stem separation and lyric editing in the browser. You can test quality on your own material before paying for credits or a subscription.
- Clear per-minute API rates: AudioShake publishes credit rates for every API model on its developer site. Teams can estimate costs in advance, such as 6 credits for a three-minute song split into two stems.
- Enterprise security: AudioShake states that it is SOC 2 compliant, and Enterprise plans add SSO, team workspaces and a dedicated SLA. That matters for labels and studios handling unreleased masters.
Cons
- Studio prices hidden until sign-up: AudioShake does not list prices for its Creator and Power User Studio plans publicly. You must create an account to see costs, and directories quote conflicting figures as a result.
- Dollar cost per credit unclear: AudioShake’s developer docs list credit rates but not the dollar price of a credit. Budgeting a large project means checking the dashboard or the AWS Marketplace listing first.
- Expensive premium models: AudioShake charges 10.0 credits per minute for multi-voice separation and music removal, ten times a standard stem. Long podcasts or films can therefore use credits quickly.
- Enterprise-first focus: AudioShake builds much of its offer around labels, studios and developers, with API contracts listed from $1,000 a month on AWS Marketplace. Hobbyists may find consumer-focused tools simpler and cheaper.
- Some features still arriving: AudioShake Studio marks music removal as coming soon in the web app, though the API already lists it. Features can therefore differ between Studio and the developer platform.
Who is Using AudioShake?
- Record labels and publishers: Create stems for remasters, remixes and sync licensing.
- Film, TV and sports media: Separate dialogue, music and effects for dubbing and highlights.
- Developers and platforms: Add stem separation to apps through the API or SDK.
- Musicians and producers: Isolate vocals or instruments for sampling, practice and mixing.
Pricing
AudioShake is freemium: Studio has a free plan, paid tiers appear after sign-up, and API usage is billed in credits per minute of audio (official site and developer docs, October 2026).
| Plan | Price | What you get |
|---|---|---|
| Studio Free | $0 | Stem separation and lyric editing in the browser |
| Studio Regular Creator | Shown after sign-up | Access to all Studio tools for regular users |
| Studio Power User | Shown after sign-up | More monthly credits and faster processing |
| Studio Enterprise | Custom | SSO, team workspaces and dedicated SLA |
| API credits | 1.0 to 10.0 credits per minute per model | Pay as you go via the dashboard; volume discounts on request |
| API via AWS Marketplace | $1,000 per month plus usage | From $0.10 per minute for speech cleanup to $2.00 per minute for instrument stems |
AIChief reports a $29 per month Pro plan, which we could not match to any plan on the official site. Treat directory figures as indicative only, since AudioShake shows Studio prices inside your account and rates can change.
What Makes AudioShake Unique?
Enterprise-grade separation across music, film and speech is what sets AudioShake apart. Many stem splitters target musicians only, while AudioShake also offers dialogue and effects separation, multi-speaker isolation, music removal for copyright clearance and lyric alignment from the same models. It backs this with a Sony Demixing Challenge win, SOC 2 compliance and on-device SDK options, which explains why labels such as Universal and Warner and media groups such as Disney appear on its customer list.
Compatibilities and Integrations
- Platforms: Browser-based Studio with no installation, plus a dashboard for API billing.
- API and SDK: REST API with quickstart guides and a real-time, low-latency SDK.
- Partner integrations: Available through services such as cielo24, OOONA, Dubverse and Algoriddim.
- File formats: Uploads up to 2GB, including WAV, FLAC, MP3, MP4 and MOV; exports up to 192kHz.
AudioShake Tutorials
A first set of stems in AudioShake Studio takes only a few steps.
- Sign up for a free account at studio.audioshake.ai.
- Choose a workflow such as Music Stems, Post Production or Speech Recovery.
- Upload your audio or video file, up to 2GB.
- Select the stems you need, such as vocals, drums or dialogue, and start processing.
- Preview the results and download the stems in your preferred format.
How We Rated It
| Criterion | Score | Why |
|---|---|---|
| Accuracy and Reliability | 4.6 | Award-winning separation quality |
| Ease of Use | 4.2 | Simple browser Studio |
| Functionality and Features | 4.7 | Music, film and speech |
| Performance and Speed | 4.1 | Fast, real-time SDK |
| Customization and Flexibility | 4.0 | Many targets, model choice |
| Data Privacy and Security | 4.4 | SOC 2, enterprise SSO |
| Support and Resources | 3.8 | Docs, FAQ, enterprise SLAs |
| Cost-Efficiency | 3.3 | Opaque Studio pricing |
| Integration Capabilities | 4.5 | API, SDK, partner apps |
| Overall Score | 4.2 | Average of the nine criteria |
Summary
AudioShake suits labels, post-production teams and developers who need high-quality stems from music, film or speech at scale. Its separation quality, breadth of models and enterprise credentials are strong, and the free Studio plan lets you judge results first. The main limitation is pricing transparency, as Studio plan costs only appear after sign-up and credit values sit in the dashboard. Process a short clip with two targets first to see how quickly credits are used.
Reviews
Write a Review