Mercury in Production

Deploy diffusion LLMs in real systems, with controls for privacy, reliability, and scale.

Enterprise-grade by design

Everything you need to evaluate, deploy, and operate Mercury with confidence.

• Security & data handling
• Quality, performance & scale
• Control, SLAs, & support
• No training on your data
• Prompts and outputs treated as customer data
• Configurable data retention & caching
• Vulnerability disclosure and tracked remediation

Deploy Mercury your way

Everything you need to evaluate, deploy, and operate Mercury with confidence.

Inception API

Fastest path to production

• Managed API with keys and usage controls
• Best for teams moving quickly

AWS Bedrock

Deploy via your existing AWS procurement and governance

• Marketplace-style buying and account-level controls
• Works with enterprise AWS environments

Azure Foundry

Standardize inside Microsoft’s AI stack

• Enterprise identity + governance alignment
• Easy adoption for Microsoft-native orgs

Model Routers

Try Mercury wherever you already route models

• Available through OpenRouter and Poe
• Integrate with a one-line code change

All deployment options can include:

• No prompt logging / no retention modes (as applicable)
• Private networking (e.g., private endpoints / VPC connectivity)
• Dedicated capacity or throughput guarantees
• Custom terms (security, legal, procurement)

We work where you already build and buy

Everything you need to evaluate, deploy, and operate Mercury with confidence.

Meet the teams who build on Mercury

Teams choose Mercury when latency and cost become product constraints

• Accelerating Low-Code Web Development with Mercury Edit Read story
• Instant Decisions for Faster and Cheaper Agentic Work Read story
• Next Edit powered by Mercury Edit Read story

Enterprise access to Mercury 2

See how fast reasoning changes your product. Benchmark on real inputs, at real scale.