specification: API Commons Plans
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans
provider: Genmo
providerId: genmo-ai
created: '2026-07-11'
modified: '2026-07-11'
reconciled: false
tags:
- Video Generation
- AI Video
- Text-to-Video
- Open Source
- Plans
description: >-
  Genmo has two distinct cost models, neither of which is a metered first-party
  developer API. First, the Mochi 1 model is open source (Apache-2.0) and free to
  self-host - you provide the GPU (~60GB VRAM) and pay only for your own compute.
  Second, the Genmo consumer web app / Playground (genmo.ai/play) is sold on a
  monthly credit system - Free, Lite, and Standard tiers - where a Mochi video
  generation costs 100 credits and a Replay costs 10 credits. Programmatic hosted
  billing is not offered by Genmo; it comes from third-party partners such as
  Replicate (approximately $0.42 per Mochi run) and fal.ai, each with their own
  pricing. The USD prices for the Lite and Standard web-app tiers did not render
  on the pricing page at capture time and are left unreconciled.
notes: >-
  Yearly billing on the web-app tiers is advertised at roughly 20% off monthly.
  Verify current credit counts, per-tier USD prices, and per-generation credit
  costs on https://www.genmo.ai/pricing during reconciliation. Third-party hosted
  rates (Replicate, fal.ai) are governed by those platforms, not Genmo.
sources:
- https://www.genmo.ai/pricing
- https://www.genmo.ai/play
- https://github.com/genmoai/mochi
- https://replicate.com/genmoai/mochi-1
plans:
- id: genmo-open-source
  name: Mochi 1 Open Source (Self-Hosted)
  type: free
  description: >-
    The full Mochi 1 text-to-video model, Apache-2.0 licensed and free to
    download from GitHub and Hugging Face for personal or commercial use. Run it
    via the repo's Python pipeline, CLI, Gradio UI, or ComfyUI. Cost is your own
    GPU compute only (single-GPU inference needs roughly 60GB VRAM).
  entries:
  - label: Self-Hosted Inference
    name: self_hosted_inference
    type: usage
    metric: videos
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: free (self-hosted; you provide GPU compute)
    userMultiplied: false
  elements:
  - name: Apache-2.0 License
  - name: Commercial Use Allowed
  - name: Python Pipeline API
  - name: ComfyUI Support
- id: genmo-web-free
  name: Web App - Free
  type: free
  description: >-
    Free tier of the Genmo consumer web app / Playground. Includes a monthly
    credit allowance and applies a Genmo watermark to outputs.
  entries:
  - label: Monthly Credits
    name: web_free_credits
    type: usage
    metric: credits
    limit: 50
    timeFrame: month
    geo: global
    unit: 1
    price: $0
    userMultiplied: false
  elements:
  - name: 50 Credits per Month
  - name: Genmo Watermark
  - name: Mochi Generation (100 credits each)
- id: genmo-web-lite
  name: Web App - Lite
  type: paid
  description: >-
    Paid consumer tier with a larger monthly credit allowance, no watermark,
    commercial usage, and higher queue priority. USD price did not render at
    capture time.
  entries:
  - label: Monthly Credits
    name: web_lite_credits
    type: usage
    metric: credits
    limit: 1200
    timeFrame: month
    geo: global
    unit: 1
    price: paid (USD price not reconciled; see pricing page)
    userMultiplied: false
  elements:
  - name: 1,200 Credits per Month
  - name: No Watermark
  - name: Commercial Usage
  - name: High Queue Priority
- id: genmo-web-standard
  name: Web App - Standard
  type: paid
  description: >-
    Top consumer tier with the largest monthly credit allowance, no watermark,
    commercial usage, highest priority queue, and early model access. USD price
    did not render at capture time.
  entries:
  - label: Monthly Credits
    name: web_standard_credits
    type: usage
    metric: credits
    limit: 5000
    timeFrame: month
    geo: global
    unit: 1
    price: paid (USD price not reconciled; see pricing page)
    userMultiplied: false
  elements:
  - name: 5,000 Credits per Month
  - name: No Watermark
  - name: Commercial Usage
  - name: Highest Priority Queue and Early Model Access
- id: genmo-third-party-hosted
  name: Third-Party Hosted (Replicate / fal.ai)
  type: usage
  description: >-
    Pay-per-run hosted inference from partners, not billed by Genmo. Replicate
    runs genmoai/mochi-1 at approximately $0.42 per run on H100 hardware; fal.ai
    hosts fal-ai/mochi-v1 (listed deprecated) with its own rates.
  entries:
  - label: Hosted Run
    name: hosted_run
    type: usage
    metric: videos
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: approximately $0.42 per run on Replicate (partner-billed)
    userMultiplied: false
  elements:
  - name: Replicate Hosted API
  - name: fal.ai Hosted API (Deprecated)
  - name: Partner-Governed Pricing
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com