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: JarvisLabs
providerId: jarvislabs
created: '2026-06-21'
modified: '2026-06-21'
reconciled: false
tags:
- AI
- GPU
- Cloud
- Infrastructure
- Compute
- Rate Limiting
- Quotas
- Throttling
description: >-
  JarvisLabs does not publish explicit numeric request rate limits for its
  instance-management API in public documentation. In practice, capacity is
  governed by real-time GPU availability (surfaced via `jl gpus`), per-account
  concurrent instance and GPU-count limits (up to 8 GPUs per instance), and
  wallet balance / grants rather than by request-per-minute throttling. Specific
  per-account values are not reconciled in this artifact.
notes: >-
  Verify any per-account concurrency caps, GPU quotas, and availability
  constraints in the JarvisLabs dashboard or with support on reconciliation;
  limits depend on account standing, region, and live capacity.
sources:
- https://docs.jarvislabs.ai/
- https://docs.jarvislabs.ai/cli
- https://jarvislabs.ai/pricing
responseCodes:
  throttled: 429
limits:
- name: Concurrent Instances
  scope: account
  metric: instances
  limit: see provider dashboard
  notes: Number of simultaneously running instances per account; not publicly numbered.
- name: GPUs Per Instance
  scope: instance
  metric: gpus
  limit: 8
  notes: Up to 8 GPUs can be attached to a single instance / deployment.
- name: GPU Availability
  scope: region
  metric: gpus
  limit: real-time
  notes: Launch success depends on live GPU availability per type and region (jl gpus).
- name: Wallet Balance
  scope: account
  metric: balance
  limit: prepaid
  notes: Usage is gated by prepaid wallet balance and any account grants.
policies:
- name: Capacity-Gated Provisioning
  description: Instance creation is constrained by real-time GPU availability per type and region rather than fixed request quotas.
- name: Backoff Strategy
  description: Clients should implement retry with backoff on capacity errors and honor any Retry-After headers.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com