About OpenGPU Radar
OpenGPU Radar is an independent AI compute intelligence platform dedicated to cost transparency across cloud GPU providers. We have no affiliate partnerships, no sponsored placements, and no vendor bias. Our mission is to provide verified, reproducible compute cost data so that developers and organizations can make informed infrastructure decisions.
Our Mission
The cloud GPU market is fragmented across dozens of providers with opaque pricing, varying spot instance behavior, and inconsistent benchmark methodologies. OpenGPU Radar exists to eliminate this information asymmetry. We aggregate, verify, and publish pricing and performance data so that every developer — from solo researchers to enterprise ML teams — has access to the same high-quality intelligence.
We do not accept paid placements, affiliate commissions, or vendor sponsorships that influence our rankings or comparisons. Every data point is independently sourced and cross-validated. When we publish a rate, it is either directly measured from our test environment, sourced from the provider's official API, or clearly labeled as an estimate derived from architectural specifications.
Data Verification Practices
Official Provider APIs
Where providers expose public APIs (Vast.ai, Spheron, Lambda Labs), we poll them on a defined cadence and record the raw response. Prices are extracted from the API response body, not from scraped HTML, ensuring reproducibility.
Community Verification
For providers without public APIs, we rely on community-submitted deployment receipts and cross-reference them against publicly documented pricing pages. All community-submitted data is flagged with the submission timestamp and verification status.
Open Feedback Loops
We maintain an open issue tracker on our GitHub repository where anyone can report incorrect pricing, outdated benchmarks, or missing models. Corrections are reviewed within 48 hours, applied to the affected pages, and preserved in our public git history for full auditability.
Open Source & Transparency
OpenGPU Radar is committed to open-source principles. Our data pipelines, benchmark scripts, and entity graph specifications are publicly available. We believe transparency in methodology is as important as transparency in pricing. Every calculation, every formula, and every data source is documented in our Methodology page.
The entity graph that powers our model, GPU, and provider pages is defined in src/lib/entity-resolver.ts and src/lib/provenance.ts. Anyone can audit the provenance chain from raw data to published page.