Lambda Research Grant (Up to $5,000)

Overview

Lambda runs a Research Grant program for academic AI/ML work. The official program page states that Lambda is "supporting groundbreaking research by offering qualifying researchers up to $5,000 in cloud credits." This is an application-based grant rather than a self-serve free tier; Lambda Cloud has no general signup credit for non-researchers.

What the Credits Cover

Grant credits apply to Lambda's on-demand GPU instances, which include NVIDIA B200, H100, A100, A6000, A10, and Quadro RTX 6000 configurations. Rates run from $0.69/hour for a Quadro RTX 6000 and $1.09/hour for an A6000 up to $3.99/hour for an H100 SXM and $6.69/hour for a B200 SXM6 (8-GPU instances, per GPU), so $5,000 covers roughly 1,250 GPU-hours on an H100 SXM. Instances ship with Lambda Stack (CUDA, cuDNN, NCCL, PyTorch, TensorFlow, JAX, Docker, JupyterLab), persistent storage filesystems are billed separately, and egress is free.

How to Apply

1. Review the featured projects on lambda.ai/research.
2. Submit the application form linked from the page's Apply for a grant button.
3. Provide your research proposal, institutional affiliation, and compute requirements.
4. Decisions are handled on a rolling basis; no deadline or decision timeline is published.

Note: a Lambda Cloud account with a verified phone number and a registered payment method is required to run instances, and grant credits are applied as service credits consumed at each weekly billing cycle.