generated: '2026-07-21'
method: searched
source: https://docs.simplismart.ai/get-started/playground + https://docs.simplismart.ai/inference/shared-endpoint
summary: >-
  Simplismart exposes an interactive Playground and shared endpoints for trying models
  before provisioning dedicated infrastructure. There is no separate test-vs-live key
  scheme or synthetic test credentials — the same organisation API key is used, and cost
  is governed by GPU quotas rather than a sandbox mode.
playground:
  url: https://app.simplismart.ai/
  docs: https://docs.simplismart.ai/get-started/playground/large-language-models
  surfaces:
    - name: Large Language Models
      docs: https://docs.simplismart.ai/get-started/playground/large-language-models
    - name: Image Generation Models
      docs: https://docs.simplismart.ai/get-started/playground/image-generation-models
    - name: Transcription Models
      docs: https://docs.simplismart.ai/get-started/playground/transcription-models
endpoint_modes:
  - mode: shared
    description: Access models through Simplismart's shared playground endpoints (no dedicated infra).
    docs: https://docs.simplismart.ai/inference/shared-endpoint
  - mode: dedicated
    description: Deploy models on exclusive, isolated infrastructure with stable performance.
    docs: https://docs.simplismart.ai/inference/dedicated-endpoint
  - mode: bring-your-own-compute
    description: Deploy and run models on your own cloud infrastructure.
    docs: https://docs.simplismart.ai/inference/bring-your-own-compute
default_quota:
  note: Every new organisation receives 1x H100 and 1x L40 GPU quota; increases via support@simplismart.ai.
  docs: https://docs.simplismart.ai/model-suite/settings/quotas
test_credentials:
  published: false
  note: No synthetic/test API keys, cards, or magic identifiers are published.