Baseten Serverless Inference Credits

Baseten provides serverless GPU inference for custom and open-source ML models. Their startup program offers credits that reduce the cost of deploying models for agents, APIs, and batch jobs.

What you get

• Serverless inference credits. Apply credits to GPU and CPU inference workloads.
• Custom model support. Deploy your own fine-tuned models, not just off-the-shelf APIs.
• Autoscaling. Scale to zero and handle traffic spikes without manual tuning.
• Production tooling. Observability, A/B testing, and model management features.

Eligibility

• Startups and small teams with a deployed or planned model inference workload.
• Credit tiers vary by application and expected volume.

How to apply

1. Visit Baseten and sign up.
2. Apply for the startup or credits program through their site.
3. Describe your model, traffic patterns, and use case.
4. Credits are applied to your account upon approval.

Caveats

• SMF Clearinghouse does not administer this offer.
• Credit terms and amounts are controlled by Baseten and may change.
• Serverless pricing still applies after credits are exhausted.

Best for

• Teams deploying custom models for agents, recommendation, or classification APIs.
• Startups that want serverless GPU without building their own platform.

Related:

• Modal Serverless GPU
• RunPod GPU Cloud