specification: API Commons Plans
specificationVersion: '0.1'
generated: '2026-07-19'
method: searched
source: https://flex.ai/pricing and https://docs.flex.ai/inference-api/reference/billing
provider: FlexAI
providerId: flexai
created: '2026-07-19'
modified: '2026-07-19'
tags:
  - Inference
  - Usage Based
  - GPU Compute
description: >-
  FlexAI is usage-priced. Token Factory (serverless) is billed per model by
  usage (per million tokens for text/vision, per image, per generated video,
  per minute of audio, per million characters for TTS). Dedicated endpoints are
  priced per GPU-hour metered per second. New accounts receive $10 free credit;
  request rate limits are tiered free (10 rpm) vs. paid (100 rpm).
plans:
  - id: flexai-token-factory-serverless
    name: Token Factory (Serverless)
    type: usage-based
    description: Pay-per-use serverless inference across open models, priced per model.
    entries:
      - label: Text / chat / reasoning input
        name: text_input
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        unit: 1000000
        price: per-model (per million input tokens)
        userMultiplied: false
      - label: Text / chat / reasoning output
        name: text_output
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        unit: 1000000
        price: per-model (per million output tokens)
        userMultiplied: false
      - label: Embeddings
        name: embeddings
        type: metered
        metric: tokens
        limit: -1
        timeFrame: usage
        unit: 1000000
        price: per-model (per million input tokens)
        userMultiplied: false
      - label: Image generation
        name: images
        type: metered
        metric: image
        limit: -1
        timeFrame: usage
        unit: 1
        price: per image
        userMultiplied: false
      - label: Video generation
        name: video
        type: metered
        metric: video
        limit: -1
        timeFrame: usage
        unit: 1
        price: per generated video
        userMultiplied: false
      - label: Speech-to-text
        name: stt
        type: metered
        metric: minute
        limit: -1
        timeFrame: usage
        unit: 1
        price: per minute
        userMultiplied: false
      - label: Text-to-speech
        name: tts
        type: metered
        metric: character
        limit: -1
        timeFrame: usage
        unit: 1000000
        price: per million characters
        userMultiplied: false
  - id: flexai-dedicated-endpoints
    name: Dedicated Endpoints
    type: usage-based
    description: Reserved GPU throughput for your models and fine-tunes, priced per GPU-hour metered per second.
    entries:
      - label: Dedicated GPU
        name: gpu_hour
        type: metered
        metric: gpu_hour
        limit: -1
        timeFrame: usage
        unit: 1
        price: per GPU-hour (metered per second)
        userMultiplied: false
  - id: flexai-ai-factory
    name: AI Factory (Private Cloud)
    type: custom
    description: FlexAI platform on your own hardware — VPC, on-prem, or air-gapped. Custom pricing; contact sales.
    entries:
      - label: Custom private-cloud tier
        name: custom
        type: custom
        metric: contract
        limit: -1
        timeFrame: usage
        unit: 1
        price: contact sales
        userMultiplied: false