Serverless inference with per-token pricing (pay for output, input, and cached tokens). Training pricing from $0.50 to $10.00 per 1M training tokens depending on model size and fine-tuning type. On-demand deployments at $0.134 to $0.334 per minute ($8 to $20 per hour) for various GPU types.
Free for one developer with up to 25 source-built services, 50,000 tool calls a month, then $0.20 per thousand. Team plan is $25 per developer per month with 100,000 tool calls per developer a month, pooled, then $0.20 per thousand. Enterprise from $1,500 per month.
Key features
Guided training runs
Configuration-led training
Custom training logic
Serverless inference
On-demand dedicated deployments
Reserved capacity
OpenAI compatible API
Model library with latest open models
Reinforcement learning support
Multi-region deployments
Source-built servers with TypeScript or Python projects
Remote MCP, OpenAPI, npm package, and container image support
Per-tool and per-argument access control policies
Tool definition review and approval workflow
Content scanning with literal and regex detectors
Activity tracking and audit trails
OIDC and SAML 2.0 single sign-on
SCIM 2.0 provisioning
Workspace-specific data encryption
Custom domain publishing with DNS and Cloudflare certification
What makes it different
Drop-in replacement for closed-model APIs with cost savings of 50-75%
Instant production deployment from any training checkpoint in seconds
Optimized inference engine for industry-leading throughput and latency
Elastic and global RL inference scaling
Own the complete learning loop for specialized intelligence
One endpoint replaces every client config, eliminating need for each client to store every server and secret
Coding agent builds the server automatically with no adapter needed
Credentials held once by Rayrun instead of distributed across clients
Built-in security review with published controls and audit trails for compliance teams
Support for multiple service types without writing adapters
Rayrun