Skip to content

LLM Licensing and Cost Management

Last reviewed: August 2026

Seat (per-user) and API (per-token) are separate billing models. Most enterprises use both.

Plan Price Audience Difference from Business
Business (formerly Team) ~$25/seat/mo Small teams (2+)
Enterprise ~$45–75/seat/mo (negotiated, 150+) Large orgs SSO/SCIM, RBAC, audit logs, EKM, data residency, custom SLA, HIPAA eligible

When to upgrade Business → Enterprise:

  • 150+ seats or security/compliance requirements (SSO mandatory, audit logs, data residency)
  • HIPAA-covered data processing
  • Org-wide governance (role-based permissions, group credit limits)

API tiers auto-upgrade based on cumulative spend. (Official Rate Limits)

Tier Condition Monthly Limit RPM/TPM Level
Free Allowed region $100/mo Low
Tier 1 $5 spent $100/mo Medium
Tier 2 $50 spent $500/mo Medium+
Tier 3 $100 spent $1,000/mo High
Tier 4 $250 spent $5,000/mo High+
Tier 5 $1,000 spent $200,000/mo Maximum
Plan Price Audience Difference from Team
Team Standard ~$25/seat/mo (official pricing) Small teams (minimum seats/caps per official page)
Team Premium ~$125/seat/mo (official pricing) High-usage teams Higher usage allowance
Enterprise Contract-based — seats + API usage, etc. (official info) Large orgs SCIM, audit logs, Compliance API, CMEK, HIPAA/BAA, org-level spend caps
Tier Level Notes
Start Entry Low RPM/TPM
Build Dev/Test Medium
Scale Production High RPM/TPM, customizable via Enterprise contract

Official Rate Limits: platform.claude.com/docs/en/api/rate-limits


Vendor Method When to Use
Azure PTU (Provisioned Throughput Unit) Fixed hourly billing, guaranteed throughput >150–200M tokens/mo steady traffic
Bedrock Provisioned Throughput Reserved model units (1/6-month commit) High-volume steady workloads + latency guarantees
Bedrock On-demand Per-token billing, no reservation Burst/experimental/irregular workloads

Channel Cost Tracking Budget/Alerts Team Allocation
OpenAI Platform Usage Dashboard Per-project monthly budget cap Projects + API Keys
Anthropic Console usage dashboard Per-workspace spend cap Workspaces
Azure Foundry Microsoft Cost Management Azure Budgets + alerts Resource tags (project, team)
Bedrock AWS Cost Explorer + CUR 2.0 AWS Budgets + Cost Anomaly Detection Inference profiles + cost allocation tags
Pattern Description
Showback Visualize team usage without actual chargeback. Awareness building
Chargeback Deduct from team budgets. Effective at preventing agent cost runaway
Model Routing Simple tasks → lightweight model; complex tasks → frontier model
Token Budgets Daily/monthly token caps per project/team/user
AI Gateway LiteLLM, Portkey, etc. for virtual key issuance, hard budgets, routing control

Seat/API pricing, minimum seats, and tier limits change frequently. Verify against the official pages below.