Pricing, limits, and billing

Look up exact InstantML plan limits, API rate limits, overage pricing, and Stripe billing behavior.

Open .md

Use this page as the authoritative reference for every plan number: included usage, rate limits, overage pricing, reset behavior, and billing states. For plan positioning and what each tier is for, see Pricing.

Plans

PlanPriceSeatsHosted storageProjectsRunsMetric points / monthAPI requests / month
Free$0 / workspace / month22 GiB21001M500k
Pro$199 / workspace / month31 TiB100100k250M25M
Premium$699 / workspace / month105 TiB5001M2B150M

Included seats apply to organization workspaces. Personal workspaces stay single-owner; create a team organization workspace to use invitations and the included seat pool.

Premium can also use customer-owned self-hosted GCP ClickHouse (BYOC) for teams that need tenant data routed to their own deployment. Customer-owned ClickHouse bytes are not counted as InstantML-hosted storage; only artifact bytes stored by InstantML are counted.

Overage

InstantML does not bill tracked training hours in v1.

Usage typeFreeProPremium
Hosted storageBlocked at included pool$0.03 / GiB-month overage$0.03 / GiB-month overage
API requestsBlocked at 500k / month$2 / 1M requests overage$1 / 1M requests overage

Paid storage overage is based on current-month high-water retained hosted storage over the included pool. API request overage is metered after the included monthly request allowance.

Paid storage overage is reportable/metered at $0.03 / GiB-month, but artifact-storage and import writes can still be blocked at plan guardrails until usage, plan, or custom terms change.

Project, run, and metric-point limits are write guardrails, not billable meters. New writes can be blocked until usage, plan, or custom terms are adjusted.

Extra seats are available through billing at $99 per seat/month.

API rate limits

Short-window API rate limits protect the hosted API and ClickHouse data planes. They apply in addition to monthly API request allowances.

PlanGeneral API rateIngest API rate
Free5 req/sec, burst 302 req/sec, burst 10
Pro50 req/sec, burst 25025 req/sec, burst 125
Premium200 req/sec, burst 1000100 req/sec, burst 500

Short-window 429 responses include X-InstantML-RateLimit-Scope: second, Retry-After, RateLimit-Limit, RateLimit-Remaining, and RateLimit-Reset.

Monthly API quota 429 responses include X-InstantML-RateLimit-Scope: monthly and reset at the next UTC calendar month. The Python SDK retries bounded short-window 429 responses and does not retry monthly quota failures.

Reset behavior

CounterReset behavior
Metric pointsCurrent UTC calendar month
API requestsCurrent UTC calendar month
ProjectsRetained resource
RunsRetained resource
SeatsRetained resource
Hosted storageRetained resource with monthly high-water overage reporting
ArtifactsVisibility-only retained count
API keysRetained resource

Settings shows the current usage period, monthly reset date, API request usage, plan rate limits, storage usage, seats, projects, runs, artifacts, and API-key counts.

Check usage from automation

Read the same counters shown in Settings with an unrestricted org API key that has usage:read:

bash
curl https://api.instantml.ai/api/usage \
  -H "Authorization: Bearer $INSTANTML_API_KEY"

The response includes the active plan catalog, current plan, limits, monthly reset window, metric-point usage, API request usage, retained storage, artifact counts, API-key counts, and billable overage fields.

Billing

Paid signup and plan changes use Stripe Checkout or Stripe Billing flows. Paid subscriptions include the base plan item plus metered overage items for hosted storage and API requests.

When an organization is pending payment or payment-failed, product write routes can return payment_required. Reads remain available unless the account state requires operator intervention.

Next steps