TrueFoundry · Rate Limits

Truefoundry Rate Limits

TrueFoundry's AI Gateway exposes plan-level monthly request quotas (50k Developer, 1M Pro/Pro Plus, 10M+ Enterprise) but does not publish a dedicated per-second / per-minute throttling document on its public site. Per-key gateway rate-limit policies are configurable by tenants for downstream models; the platform-side request budget is enforced as a monthly quota.

6 Limits Throttle: 429
AI GatewayLLMOpsGenAIRate Limiting

Limits

Developer monthly request quota account
requests_per_month · month
50000
Free Developer tier monthly cap.
Pro monthly included requests account
requests_per_month · month
1000000
Included with the $499/month Pro subscription; overages billed.
Pro Plus monthly included requests account
requests_per_month · month
1000000
Included with the $2,999/month Pro Plus subscription.
Enterprise monthly request floor account
requests_per_month · month
10000000
Starting allotment; negotiated upward for enterprise contracts.
Pro Plus MCP tool-call quota account
calls_per_month · month
5000000
MCP gateway tool-call quota included with Pro Plus.
Tenant-configurable gateway throttle virtual-key
varies
tenant-defined
Customers can configure RPM / TPM limits per virtual key, model, or team in the AI Gateway control center.

Policies

Plan Quota Enforcement
Monthly request quotas are enforced at the account level; overage on Pro is billed per vendor pricing, while Developer and Pro Plus appear capped at their published numbers.
Tenant Self-Service Throttling
Tenants configure their own per-key, per-model, or per-team rate limits in the gateway control plane; these supplement the platform monthly quotas.
Standard 429 Retry
Exceeding gateway-side limits returns HTTP 429; clients should implement exponential backoff with jitter and respect any Retry-After header.

Sources