specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Google Colab
providerId: google-colab
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-05'
reconciled: true
tags:
  - Notebooks
  - Machine Learning
  - Google Cloud
  - Rate Limiting
description: Colab does not expose a request-rate API; instead it enforces session-level limits on runtime
  duration, idle disconnect, concurrent sessions, and GPU/TPU availability. Resource access is governed
  by your subscription tier and remaining compute units. Colab Enterprise inherits the standard Vertex
  AI / Notebooks API quotas.
sources:
  - https://research.google.com/colaboratory/faq.html
  - https://colab.research.google.com/signup
  - https://cloud.google.com/colab/pricing
responseCodes:
  resourceExhausted: 429
  unavailable: 503
limits:
  - name: Maximum runtime duration (Free)
    scope: session
    metric: hours
    limit: 12
    timeFrame: usage
    notes: Free notebooks disconnect at 12 hours of continuous runtime; idle disconnects much earlier.
  - name: Idle timeout (Free)
    scope: session
    metric: minutes
    limit: 90
    timeFrame: usage
    notes: Approximate; Colab disconnects idle Free sessions sooner than paid tiers.
  - name: Maximum runtime duration (Pro)
    scope: session
    metric: hours
    limit: 24
    timeFrame: usage
  - name: Maximum runtime duration (Pro+)
    scope: session
    metric: hours
    limit: 24
    timeFrame: usage
    notes: Pro+ also supports persistent background execution beyond the active session window.
  - name: Concurrent sessions per user
    scope: user
    metric: sessions
    limit: 1
    notes: Free and Pro typically support 1 active high-resource session; Pro+ allows additional concurrent
      sessions.
  - name: GPU/TPU availability
    scope: user
    metric: best-effort
    limit: see availability
    notes: Premium GPUs (A100/H100/L4) are subject to availability and prioritized for paid tiers.
  - name: Compute unit consumption
    scope: subscription
    metric: compute_unit_per_session
    limit: see live runtime meter
    notes: Compute units are consumed per hour of attached runtime; rate varies with GPU/TPU class.
policies:
  - name: Session lifecycle
    description: Sessions terminate on idle disconnect, runtime cap, or browser close. Save progress to
      Drive or GitHub frequently.
  - name: Resource gating by tier
    description: Free tier is best-effort and may queue or refuse high-end GPUs during peak load. Paid
      tiers receive priority but are still capacity-limited at the global pool.
  - name: Compute unit metering
    description: Pro/Pro+ subscribers consume from a monthly compute-unit bucket; PAYG users consume from
      a 90-day bucket. Once exhausted, runtimes downgrade to Free behavior.
  - name: Enterprise quotas
    description: Colab Enterprise (Vertex AI Workbench) follows standard Vertex AI / Compute Engine quotas
      enforced at the GCP project level; raisable via the Cloud Console quotas page.
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com