Home / Alternatives / Vast.ai for Transcription
Vast.ai Alternatives for Transcription
Speech to text over a batch of audio, usually Whisper, where the run is long but each file is independent.
What transcription actually needs
Why people look past Vast.ai for this
Per-file independence makes this the easiest job to restart elsewhere; the only real loss is the files already processed if outputs were not written out as they completed. That is exactly the gap Vast.ai does not close: Vast.ai is usually the cheapest raw dollar-per-hour on the market, because hosts bid against each other and you absorb the variance in reliability. The honest framing is not us-versus-them on price: it is whether the environment you build survives the box it was built on.
To be fair, Vast.ai's real strength is real: Raw price. A host-bid marketplace with interruptible instances will beat brokered on-demand capacity on the low end, and that is by design.
Live rates for the GPUs transcription wants
| GPU | From (per GPU/hr) |
|---|---|
| L40S | $0.869 |
| RTX 4090 | $0.374 |
| A100 | $1.21 |
| L4 | $0.649 |
Live per-GPU rates from Aquanode's marketplace. Refreshes hourly.
How you run it here today
Run it today with the seeded whisper-transcribe job recipe: a real, publicly-imaged container Aquanode can queue directly, not a template we're promising to build later.
Restoring an environment requires a snapshot that already exists. Stopping a deployment yourself captures it on the way out, so you can bring it back later on any provider. A provider-side termination is different: it is only recoverable if you had already switched automated snapshots on for that deployment, and it costs you the work since the last one. Automated snapshots are opt-in, nothing runs until you start it, and with none running there is nothing to restore.
More alternatives pages
Other workloads on Vast.ai
Transcription alternatives to other clouds
Related reading: H100 pricing and specs, H100 vs H200, A100 vs H100, H100 and H200: SXM vs NVL vs PCIe, and The best GPUs for AI, ranked.
