---
title: "Luma Dream Machine Ray3 | Ethical Stack"
url: https://ethicalstack.ai/products/luma-ray3/
catalog_version: f6baff2119d0e9158520da7fbe4232011dac79f25bd08757436d7a1dbd593db2
---

# Luma Dream Machine Ray3

Cloud video generation and editing with Ray3; user content and commercial rights vary by plan and account terms.

Canonical profile: https://ethicalstack.ai/products/luma-ray3/

Scope: Luma Dream Machine Ray3: cloud access. Free tier; paid Dream Machine plans and credit costs depend on model, duration and resolution. Ratings describe the stated product, company or configuration in each finding.

Published overall: 3.8; 6/6 dimensions rated.

Round the equal-weight mean of available numerical ratings to one decimal; require at least four of six for an overall score. BDS status stays outside the mean; for eligible profiles, a boycott target caps the final score at 2/10 and a pressure target subtracts one point (minimum 1). Not listed and not checked do not adjust the score.

## Energy practices

Finding: 4/10

Assessment: https://ethicalstack.ai/products/luma-ray3/#energy

Luma documents Ray3 Draft Mode, model/output controls, and video length and resolution. Its training summary does not measure electricity, emissions, or water for the model or a generation.

Reason: 1 + 1 reporting + 0 electricity + 2 controls = 4/10. This grade covers documented practices and transparency; it does not measure the energy or emissions of a request.

Argument: Luma documents Ray3 Draft Mode, model/output controls, and video length and resolution. Its training summary does not measure electricity, emissions, or water for the model or a generation. The score describes documented measurement transparency and available controls, not environmental performance.

Type: conditional; confidence: medium

Scope (dimension): Ray3 generation in Luma Dream Machine web service; no local host assumed.

Research reviewed: 2026-10-24; evidence date: 2026-10-24

Unknowns: No Luma company electricity-sourcing or Ray3 service-level footprint was found. Draft mode’s speed/credit claims are not physical-resource measurements, and the local/hosted inference electricity boundary is not published for requests.

```json
{
  "method": "practices-transparency-v1",
  "scope": "Ray3 generation in Luma Dream Machine web service; no local host assumed.",
  "components": {
    "disclosure": {
      "score": 1,
      "reason": "Luma identifies Ray model families and training-content scope and documents the hosted product, but publishes no quantitative energy/carbon/water report.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/learning-hub/ray3-user-guide"
      ]
    },
    "electricity": {
      "score": 0,
      "reason": "No electricity sourcing is established for Dream Machine Ray3 generations. Luma’s training summary and product pages do not specify clean-power sourcing for a generation route.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/ray3"
      ]
    },
    "demand": {
      "score": 2,
      "reason": "Ray3 offers Draft versus full-fidelity generation, 5- or 10-second lengths for relevant workflows, 540p/720p/1080p outputs, and HDR/upscale/extension choices. These are multiple output and model controls; advertised speed/credit savings are not counted as measured physical reductions.",
      "sourceUrls": [
        "https://lumalabs.ai/learning-hub/ray3-user-guide",
        "https://lumalabs.ai/learning-hub/ray3-faqs"
      ]
    }
  },
  "actualImpactMeasured": false,
  "meaning": "Documented environmental practices and transparency. Zero component points mean a reviewed evidence gap."
}
```

- [Public Summary of Training Content: Ray Models — Luma](https://lumalabs.ai/legal/training-content-summary-ray) (primary; checked 2026-10-24)

- [Ray3: User Guide](https://lumalabs.ai/learning-hub/ray3-user-guide) (official product documentation; checked 2026-10-24)

- [Ray3 FAQ](https://lumalabs.ai/learning-hub/ray3-faqs) (official product documentation; checked 2026-10-24)

- [Ray3 video model](https://lumalabs.ai/ray3) (primary; checked not recorded)

## Training data

Finding: 4/10

Assessment: https://ethicalstack.ai/products/luma-ray3/#training-rights

Luma’s Ray summary identifies broad training categories, transactional licensing in text, image and video, a selected-site crawl, and synthetic data. It says Ray 3 began from third-party open weights but expressly leaves that base model’s original training data outside the summary.

Reason: 1 + 1 source transparency + 1 permission + 1 creator controls = 4/10. This grades documented training-data safeguards, not a legal ruling.

Argument: The source description remains at broad categories: its large public datasets are listed as N/A, its crawled domains and license partners are unnamed, and the inherited base corpus is unaccounted for. Limited licenses plus a narrow complaint/crawler process do not establish rights across the whole lineage, so it earns 1+1+1+1=4.

Type: argument; confidence: medium

Scope (dimension): The full Ray 3, Ray 3.14 and Ray 3.2 model lineage covered by Luma’s October 2026 summary, including the third-party base model. Luma’s account does not identify the base corpus; that unaccounted lineage receives no rights credit.

Research reviewed: 2026-10-24; evidence date: 2026-10-23

Unknowns: Original training sources and rights for the inherited third-party base, dataset and domain identities, proportions by source/permission, compatibility checks, and whether complaints result in removal or retraining remain undisclosed.

```json
{
  "method": "training-rights-evidence-v1",
  "scope": "The full Ray 3, Ray 3.14 and Ray 3.2 model lineage covered by Luma’s October 2026 summary, including the third-party base model. Luma’s account does not identify the base corpus; that unaccounted lineage receives no rights credit.",
  "components": {
    "provenance": {
      "score": 1,
      "reason": "The EU summary names broad source classes and Ray lineage, but reports no large public datasets, no names for licensed partners, and no actual crawled domains. Its inherited third-party base is outside the source accounting and earns no provenance credit.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray"
      ]
    },
    "permissions": {
      "score": 1,
      "reason": "Luma reports some transactional licenses across text, image and video, but also public and crawled sources without item-level permissions or proportions; the inherited base model’s rights basis is not disclosed and earns no rights credit.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray"
      ]
    },
    "creatorControls": {
      "score": 1,
      "reason": "Luma describes a targeted crawler that excludes known piracy sites and respects technical barriers, plus a rights-holder complaint route; it gives no evidence of source-level opt-out coverage or retraining after a complaint.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray"
      ]
    }
  },
  "routeDependency": null,
  "modelReference": null,
  "meaning": "Documented training-data rights safeguards. Low points may reflect missing evidence; they do not establish infringement."
}
```

- [Public Summary of Training Content: Ray Models — Luma](https://lumalabs.ai/legal/training-content-summary-ray) (primary; checked 2026-10-24)

## Your data & privacy

Finding: 3/10

Assessment: https://ethicalstack.ai/products/luma-ray3/#privacy

Luma says it collects prompts and uploaded files and uses collected information to train and improve its models; its licensing guide distinguishes free/lite from paid-plan content rights.

Reason: This is a cloud service with content collection and model-improvement uses described in the privacy policy. Enterprise agreements can set a separate processor relationship, so this score does not cover enterprise terms.

Argument: This is a cloud service with content collection and model-improvement uses described in the privacy policy. Enterprise agreements can set a separate processor relationship, so this score does not cover enterprise terms.

Type: conditional; confidence: medium

Scope (dimension): Selected configuration

Research reviewed: not recorded; evidence date: not recorded

Unknowns: This is a cloud service with content collection and model-improvement uses described in the privacy policy. Enterprise agreements can set a separate processor relationship, so this score does not cover enterprise terms.

- [Luma Privacy Policy](https://lumalabs.ai/legal/privacy-policy) (primary; checked 2026-10-24)

- [Dream Machine Guide: Licensing](https://lumalabs.ai/learning-hub/licensing) (primary; checked 2026-10-24)

## Openness

Finding: 2/10

Assessment: https://ethicalstack.ai/products/luma-ray3/#openness

Dream Machine Ray3 is offered as Luma’s hosted generative service. Luma’s terms reserve proprietary service and model rights, and its API terms prohibit attempts to access underlying models, weights, source code or training data.

Reason: Published documentation and APIs expose some ways to use the service, but not the model internals or self-hosting rights. The restrictions place Ray3 in the low openness range.

Argument: Published documentation and APIs expose some ways to use the service, but not the model internals or self-hosting rights. The restrictions place Ray3 in the low openness range.

Type: evidence; confidence: medium

Scope (dimension): Product

Research reviewed: not recorded; evidence date: 2026-05-14

Unknowns: Published documentation and APIs expose some ways to use the service, but not the model internals or self-hosting rights. The restrictions place Ray3 in the low openness range.

- [Terms of Service — Luma](https://lumalabs.ai/legal/terms-of-service) (primary; checked 2026-10-24)

- [API Terms of Use — Luma](https://lumalabs.ai/legal/api-terms-of-use) (primary; checked 2026-10-24)

## Worker protections

Finding: 2/10

Assessment: https://ethicalstack.ai/products/luma-ray3/#labour

Luma’s careers page describes employee benefits and compensation language, but no worker labor standards, contractor/supplier coverage or independently corroborated outcomes were found.

Reason: 1 + 1 pay & conditions + 0 worker voice & remedy + 0 evidence & follow-through = 2/10. This grades documented worker safeguards, not a measured rate of mistreatment.

Argument: Luma’s careers page reports 250+ team members and describes competitive compensation, medical/dental/vision coverage, flexible time off and a $1,500 annual learning stipend. This is limited employee-facing benefits information, not a labor-rights code, supplier standard or independently verified worker outcome. Luma’s model training-content disclosure does not establish the conditions of upstream data workers, annotators or contractors.

Type: argument; confidence: low

Scope (dimension): Luma AI's workforce and disclosed contractors. Coverage of individual product teams may be incomplete.

Research reviewed: 2026-10-24; evidence date: not recorded

Unknowns: Pay ranges and hours, grievance/representation mechanisms, contractor and vendor coverage, data-labeling or moderation work, health and safety, and independent worker outcomes remain undisclosed in the sources reviewed.

```json
{
  "method": "worker-safeguards-evidence-v1",
  "entity": "luma",
  "company": "Luma AI",
  "scope": "Luma AI's workforce and disclosed contractors. Coverage of individual product teams may be incomplete.",
  "components": {
    "conditions": {
      "score": 1,
      "reason": "The careers page describes employee compensation/benefits and flexibility, but provides no worker-level results or contractor/supplier conditions.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/legal/terms-of-service",
        "https://lumalabs.ai/legal/enterprise-terms-of-service",
        "https://lumalabs.ai/careers"
      ]
    },
    "voice": {
      "score": 0,
      "reason": "No reviewed Luma source establishes employee representation, grievance routes, anti-retaliation process or documented worker remedies.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/legal/terms-of-service",
        "https://lumalabs.ai/legal/enterprise-terms-of-service",
        "https://lumalabs.ai/careers"
      ]
    },
    "assurance": {
      "score": 0,
      "reason": "No independent labor audit or corroborated employee, contractor or supplier outcomes were disclosed; a model training-content summary is not labor assurance.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/legal/terms-of-service",
        "https://lumalabs.ai/legal/enterprise-terms-of-service",
        "https://lumalabs.ai/careers"
      ]
    }
  },
  "disclosureLimited": true,
  "meaning": "This grades documented worker safeguards, not a measured rate of mistreatment."
}
```

- [Public Summary of Training Content: Ray Models — Luma](https://lumalabs.ai/legal/training-content-summary-ray) (primary; checked 2026-10-24)

- [Terms of Service — Luma](https://lumalabs.ai/legal/terms-of-service) (primary; checked 2026-10-24)

- [Luma Enterprise Terms of Service](https://lumalabs.ai/legal/enterprise-terms-of-service) (terms; checked 2026-10-24)

- [Luma Careers and Benefits](https://lumalabs.ai/careers) (careers; checked 2026-10-24)

## Military ties

Finding: 8/10

Assessment: https://ethicalstack.ai/products/luma-ray3/#military-contracts

Luma’s enterprise AUP prohibits promoting or inciting violence, harmful or illegal use, and certain high-risk activities; government terms expressly contemplate civilian and military users, but no actual deployment was established.

Reason: 7 base + 1 safeguard adjustment = 8/10 (3 of 9 safeguard points; every 3 adds 1). No military relationship was established in the cited sources. This ranks above known defense ties without claiming a verified clean record.

Argument: Luma’s Enterprise Terms prohibit use that risks harm, unlawful regulated activity, violent promotion/incitement and defined high-risk activities, and allow Luma to screen/remove content. They also explicitly contemplate U.S. government use for civilian and military purposes under FAR/DFARS terms and prohibit ITAR-controlled inputs. The government clause is availability, not evidence of a named customer; the AUP does not ban military use.

Type: argument; confidence: low

Scope (dimension): Luma AI's relationships and safeguards. A company relationship does not identify every product or release used.

Research reviewed: 2026-10-24; evidence date: not recorded

Unknowns: Unpublished contracts, reseller/subcontractor activity, customer-configured use and enforcement outcomes remain unknown. Hosted AUPs do not necessarily apply to independent local deployments or permissively licensed copies.

```json
{
  "method": "military-ties-priority-v2",
  "entity": "luma",
  "company": "Luma AI",
  "scope": "Luma AI's relationships and safeguards. A company relationship does not identify every product or release used.",
  "components": {
    "restrictions": {
      "score": 1,
      "reason": "Hosted enterprise rules restrict violence promotion/incitement, illegal regulated goods and defined high-risk activities, but explicitly allow government licensees subject to FAR/DFARS and do not ban military or intelligence purposes.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/legal/terms-of-service",
        "https://lumalabs.ai/legal/enterprise-terms-of-service"
      ]
    },
    "oversight": {
      "score": 1,
      "reason": "Luma reserves monitoring, blocking/removal and service access restrictions for AUP violations, but says it has no obligation to monitor and publishes no military-specific oversight outcomes.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/legal/terms-of-service",
        "https://lumalabs.ai/legal/enterprise-terms-of-service"
      ]
    },
    "transparency": {
      "score": 1,
      "reason": "Terms transparently identify possible military government users and AUP scope, but no actual military customer, purpose or deployment is disclosed.",
      "sourceUrls": [
        "https://lumalabs.ai/legal/training-content-summary-ray",
        "https://lumalabs.ai/legal/terms-of-service",
        "https://lumalabs.ai/legal/enterprise-terms-of-service"
      ]
    }
  },
  "relationship": "not-established",
  "relationshipReason": "Luma’s terms provide government licensing language for both civilian and military end users, but the clause is boilerplate availability and no actual government customer, contract or deployment was established in targeted official searches.",
  "score": 8,
  "baseScore": 7,
  "safeguardPoints": 3,
  "safeguardAdjustment": 1,
  "rawScore": 8,
  "cap": null,
  "disclosureLimited": false,
  "meaning": "Known military relationships take priority over missing safeguards. The rating does not establish the probability of an undisclosed contract."
}
```

- [Public Summary of Training Content: Ray Models — Luma](https://lumalabs.ai/legal/training-content-summary-ray) (primary; checked 2026-10-24)

- [Terms of Service — Luma](https://lumalabs.ai/legal/terms-of-service) (primary; checked 2026-10-24)

- [Luma Enterprise Terms of Service](https://lumalabs.ai/legal/enterprise-terms-of-service) (terms; checked 2026-10-24)

## BDS targets

Finding: Not listed

Assessment: https://ethicalstack.ai/products/luma-ray3/#israel-palestine

Categorical check of the BNC campaign guide and No Tech campaign page; absence from these named-target pages is not BDS approval.

Reason: No named target for this company was found in the two checked BNC sources as of 2026-10-24. This is a narrow list check, not BDS approval, and it does not capture unlisted relationships or partners.

Argument: No named target for this company was found in the two checked BNC sources as of 2026-10-24. This is a narrow list check, not BDS approval, and it does not capture unlisted relationships or partners.

Type: conditional; confidence: medium

Scope (dimension): Company record

Research reviewed: not recorded; evidence date: not recorded

Unknowns: No named target for this company was found in the two checked BNC sources as of 2026-10-24. This is a narrow list check, not BDS approval, and it does not capture unlisted relationships or partners.

- [Guide to BDS Boycott & Pressure Corporate Priority Targeting](https://bdsmovement.net/Guide-to-BDS-Boycott) (primary; checked not recorded)

- [No Tech for Oppression, Apartheid or Genocide](https://bdsmovement.net/no-tech-oppression-apartheid-or-genocide) (primary; checked not recorded)

## Reading the evidence

- Assessments are provisional editorial judgments. Cite the profile and the underlying sources, and retain confidence, scope and unknowns.

- Unknown, not-applicable and model-dependent are different states. Missing evidence does not establish good or bad conduct.

- Company-level evidence does not establish use of a particular product or model. Account plan, model, host and route can change an assessment.

- Energy scores describe documented practices and transparency, not measured request emissions. Zero component points can mean an evidence gap.

- Published overall scores require at least four of six numerical dimensions. Preserve coverage and BDS adjustments; BDS itself is categorical.

- No named BDS target found is limited to the reviewed campaign sources and is not approval. Personalized priorities do not change the published overall score.

- Review and source-check dates describe the research, not this deployment. Fetch current primary sources when the decision requires current verification.

Methodology: https://ethicalstack.ai/methodology/