Rate-limit info
GET /v1/organizations/rate_limits returns the rate limits that apply to your organization: your tier, each model’s requests-per-minute and token-per-minute limits, and any lower limits set on individual workspaces. You can also set or clear those workspace limits.
Use it to configure client-side throttling from the real numbers instead of hard-coding them, and to check what a tier upgrade changed.
Endpoints
| Method | Path | Purpose |
|---|---|---|
GET | /v1/organizations/rate_limits | Organization tier and per-model limits. |
GET | /v1/organizations/workspaces/{workspace_id}/rate_limits | The effective limits for one workspace. |
PUT | /v1/organizations/workspaces/{workspace_id}/rate_limits | Set per-model caps for the workspace. |
DELETE | /v1/organizations/workspaces/{workspace_id}/rate_limits | Remove all workspace caps. |
Authenticate with an admin key and send syntic-beta: admin-2026-11-01.
Organization limits
{
"type": "rate_limits",
"tier": "2",
"next_tier": { "tier": "3", "paid_to_date_required_usd": "250.00", "paid_to_date_usd": "131.40", "eligible_after": "2026-11-12T00:00:00Z" },
"data": [
{
"model": "Amara S1.2",
"requests_per_minute": 1000,
"input_tokens_per_minute": 400000,
"output_tokens_per_minute": 80000
},
{
"model": "Amara Quick",
"requests_per_minute": 1500,
"input_tokens_per_minute": 800000,
"output_tokens_per_minute": 160000
}
],
"other_limits": [
{ "group": "token_counting", "requests_per_minute": 2000 },
{ "group": "files_batches_models", "requests_per_minute": 1000 },
{ "group": "admin_api", "requests_per_minute": 300 }
]
}tier is start, 1 to 4, or custom. next_tier is null at tier 4 and for custom tiers.
Workspace limits
The response for a workspace has the same data list with each model’s effective limits (the lower of the organization’s and the workspace’s) plus overrides showing what was set on the workspace:
{
"type": "workspace_rate_limits",
"workspace_id": "wrkspc_01JC5...",
"data": [{ "model": "Amara S1.2", "requests_per_minute": 100, "input_tokens_per_minute": 50000, "output_tokens_per_minute": 80000 }],
"overrides": [{ "model": "Amara S1.2", "requests_per_minute": 100, "input_tokens_per_minute": 50000 }]
}To set caps, PUT an overrides list. Each entry names a model and any of the three limits; omitted limits follow the organization’s. The list replaces previous overrides.
# Keep the staging workspace to 100 RPM on the default model
curl -sS -X PUT https://app.syntic.ai/v1/organizations/workspaces/wrkspc_01JC5B7D9F1H3K5M7P9R/rate_limits \
-H "Authorization: Bearer $SYNTIC_ADMIN_KEY" \
-H "Content-Type: application/json" \
-H "syntic-beta: admin-2026-11-01" \
-d '{"overrides": [{"model": "Amara S1.2", "requests_per_minute": 100, "input_tokens_per_minute": 50000}]}'Errors
Errors use the error envelope.
| Status | type | code | When |
|---|---|---|---|
| 400 | invalid_request_error | override_above_org_limit | A workspace cap higher than the organization’s limit. |
| 400 | invalid_request_error | unknown_model | Model name not in the catalogue. |
| 404 | not_found_error | workspace_not_found | No such workspace. |
Build notes for the platform team
- Read limits from the same tier table the limiter enforces; never keep a second copy.
- Overrides are stored per workspace and model and loaded into the limiter’s bucket configuration within 60 seconds of a change.