Home / Alternatives / Shadeform for Training
Shadeform Alternatives for Training
Training a model from scratch or continuing a pre-training run, usually for days rather than hours.
What training actually needs
Why people look past Shadeform for this
The checkpoint is the cheap part to move; the built environment, the dataset cache and the exact CUDA and framework versions around it are what take a day to rebuild. That is exactly the gap Shadeform does not close: Shadeform solves the same first half of the problem we do: one API across many GPU clouds instead of an account per vendor. Where we diverge is the second half: finding capacity is not the same as being able to take your environment with you when you move to it.
To be fair, Shadeform's real strength is real: Broad first-party cloud coverage with a single API, and a clean deployment story for teams that mainly need capacity aggregation.
Live rates for the GPUs training wants
| GPU | From (per GPU/hr) |
|---|---|
| H100 | $2.78 |
| H200 | $4.91 |
| B200 | $6.67 |
| A100 | $1.50 |
Live per-GPU rates from Aquanode's marketplace. Refreshes every 6 hours.
How you run it here today
Run it today as a Pod: the console's own box workload card preselects the right template for this job.
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 Shadeform
Training 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.
