Pricing for personal builders and enterprise teams

Personal accounts can start instantly with usage-based billing in USD. Enterprise teams can review rollout, billing, and migration needs with sales before switching traffic.

Personal

Start fast and pay as you go

Best for individual developers, prototypes, and small products that want access to the currently supported OpenAI-compatible endpoints.

Usage-based

No contract. Top up balance in USD only when you need more credits.

  • Self-serve sign up and top-ups
  • Per-key quotas and rate limits
  • Usage dashboard and billing history
  • OpenAI-compatible access for supported endpoints
Enterprise

Plan onboarding and billing with sales

Best for product teams that need rollout planning, billing review, or help moving existing traffic to Haze.

Talk to sales

We'll align onboarding steps, protocol scope, and billing review before rollout.

  • Rollout planning for team access
  • Billing review and finance coordination
  • Migration planning for existing API traffic
  • Priority onboarding through sales

Metered pricing for supported OpenAI endpoints

The model rates below apply to personal accounts today and guide enterprise rollout planning. Usage is billed per 1 million tokens in USD.

Supported protocol scope

  • GET /v1/models
  • POST /v1/chat/completions
  • POST /v1/responses
  • POST /v1/embeddings

Anthropic-compatible routing is also available for Claude tooling, but OpenAI support is limited to the endpoints listed above.

Anthropic

ModelInput/ 1M tokensOutput/ 1M tokens
claude-haiku$0.2500$1.25
claude-opus$15.00$75.00
claude-sonnet$3.00$15.00

OpenAI

ModelInput/ 1M tokensOutput/ 1M tokens
codex$2.50$10.00
gpt-3.5-turbo$0.5000$1.50
gpt-4-turbo$10.00$30.00
gpt-4o$2.50$10.00
gpt-4o-mini$0.1500$0.6000
o1$15.00$60.00
o1-mini$3.00$12.00
o3$15.00$60.00
o3-mini$1.10$4.40
Note: All prices are in USD. Includes relay infrastructure, optimized routing, and regional availability.

Need help choosing the right path?

Personal accounts are ready for self-serve launch. Enterprise teams can use the support page to ask about rollout planning, support channels, and protocol coverage before they migrate.