Policy Centre

AI Provider Register

A production-grounded register of AI providers and models used, tested, configured, or awaiting deployment in Oryyn Creative Lab.

2
Operational providers
2
Tested, configured, or awaiting deployment
provider-register-2026-07-24-v1
Register version
Availability is derived from internal catalog readiness, successful production generation evidence, and separately identified administrative API tests. Provider terms, models, regions, and policies can change.

Operational providers

Providers with successful production evidence or a ready production endpoint.

Base44 Built-in AI

Operational

Wix.com Ltd. (operator of Base44-branded services)

Availability: Successful production evidence or a ready production endpoint exists.

Models and capabilities

Base44 Built-in
Operational
Image
Base44 Built-in
Operational
Video
Base44 Built-in
Operational
Audio
base44_builtin_image
Operational
Image
base44_builtin_video
Operational
Video
base44_video
Operational
Video
Data sent
Prompts, uploaded or referenced media, generation settings, and service identifiers needed to provide built-in AI generation. Base44 may use third-party AI service providers.
Processing region
Base44 does not publicly fix one processing region for every built-in AI request. Processing may involve Base44, cloud/CDN infrastructure, and listed third-party AI providers.
Retention
Retention depends on Base44 service, account, security, backup, and legal requirements. Exact per-generation retention is not reported by the built-in generation response.
Training use
Base44 terms permit use of Customer Data and Generated Output to maintain and improve services, including training software tools, subject to the applicable agreement and privacy terms.
Commercial use
Users remain responsible for rights, lawful use, and any third-party AI terms. Generated output may not be unique and Base44 does not guarantee exclusive intellectual-property rights.
Watermark and provenance
The built-in generation response does not consistently report watermark or content-credential metadata. Creative Lab records this as not reported unless the provider supplies evidence.
Register notes
The precise underlying third-party model may not always be disclosed by the built-in generation response.
Last verified
24 Jul 2026 · Transparency status: limited

Seevio.ai

Operational

Seevio.ai (public policy does not state a separate incorporated operator name)

Availability: Successful production evidence or a ready production endpoint exists.

Models and capabilities

seedance-2-0
Operational
Video
seedance-2-0-fast
Operational
Video
Data sent
Text prompts, generation settings, and image, video, or audio references supplied for the selected video-generation mode, together with task and callback identifiers.
Processing region
The public API and privacy documentation do not state a fixed processing or storage region. Verify contractual data-location requirements before sensitive use.
Retention
The privacy policy states that data is retained only as long as necessary to provide services, without publishing a fixed per-generation retention period.
Training use
The privacy policy states that prompts, uploads, and generated videos are not used to train or fine-tune AI models.
Commercial use
The terms state that users retain uploaded-content rights and may use generated videos personally or commercially, subject to applicable law, rights ownership, and the provider terms.
Watermark and provenance
The provider website states that generated videos are watermark-free. The API does not currently return a content-credentials attestation to Creative Lab.
Register notes
Seedance2.ai migrated to Seevio.ai (same account, credits, and API per the provider's own migration notice). Creative Lab identifies the current API provider as Seevio.ai and does not represent it as a direct ByteDance-operated integration without contractual evidence. Internal key retained as seedance2 for ledger/backfill continuity.
Last verified
24 Jul 2026 · Transparency status: limited

Configured and planned integrations

These entries are not represented as generally available production generation options.

Alibaba Cloud Model Studio / Qwen

Tested

Alibaba Cloud (relevant contracting entity under the International Website Product Terms)

Availability: An administrative test succeeded; normal production availability is not confirmed.

Models and capabilities

builtin-generateimage
Tested
Image
Data sent
Prompts, source images or other enabled inputs, generation settings, workspace and request identifiers required by the selected Model Studio API.
Processing region
The selected access region and service deployment scope determine request routing and storage. Creative Lab must record the configured region before marking the provider operational.
Retention
Alibaba Cloud documents workspace and regional handling but no single public retention period applies to every inference interface. The configured service and product terms control.
Training use
Alibaba Cloud Model Studio states that customer data is not used for model training without explicit consent.
Commercial use
Commercial use remains subject to Model Studio product terms, model-specific terms, applicable law, and the user’s rights in source material.
Watermark and provenance
Watermark and content-credential behaviour is model-specific and must be recorded as not reported unless returned by the selected Qwen image endpoint.
Register notes
Current Qwen image catalog entries are configured but not production-verified at the start of this phase.
Last verified
24 Jul 2026 · Transparency status: verified

NVIDIA

Tested

NVIDIA Corporation

Availability: An administrative test succeeded; normal production availability is not confirmed.

Models and capabilities

Llama Test
Tested
Image
Data sent
Prompts, media inputs, generation or inference settings, and account/request identifiers required by the selected NVIDIA service or model endpoint.
Processing region
Region and deployment depend on the selected hosted service, cloud agreement, or self-hosted NIM deployment. Self-hosted entries do not send content to NVIDIA unless separately configured.
Retention
Retention depends on the applicable NVIDIA service-specific agreement. General terms do not establish one retention period for every API Catalog or NIM service.
Training use
General Technology Access Terms permit use of User Content to provide, secure, support, modify, and improve NVIDIA products or services. A service-specific product or cloud agreement may provide narrower terms.
Commercial use
Model and service-specific licenses govern commercial use. Catalog availability does not itself grant redistribution or production rights.
Watermark and provenance
Watermark and content-credential behaviour is model-specific and is recorded as not reported unless an endpoint supplies an attestation.
Register notes
Configured or downloadable NVIDIA models are not represented as operational until a production endpoint is ready or successful production evidence exists.
Last verified
24 Jul 2026 · Transparency status: limited
Provider and Model Policy

Third-Party AI Provider Notice

Effective date: 23 July 2026

Creative Lab integrates third-party models and services for image, video, audio, voice, enhancement, moderation, hosting, delivery and other operations.

1. Provider information

The provider or model label shown in Creative Lab at the time of generation is the current transaction-level disclosure. This page provides the general rules that apply to integrated providers. Provider availability, terms, processing locations, retention positions and commercial-use restrictions may change; users should review the provider information shown before generation and contact canute@maveristic.com if they need clarification before submitting sensitive or commercially important material.

2. Pass-through rules

When provider terms are displayed or linked before use, the user agrees to comply with them. If a provider rule and Creative Lab rule differ, the stricter safety or rights restriction applies to that provider.

3. Routing

Creative Lab may route or reroute a job based on user selection, availability, geography, safety, cost, quality, latency or compatibility. A materially different provider that changes data use or output rights should be disclosed or require user choice where practicable.

4. Data handling

A provider may receive prompts, media, settings and operational identifiers and may process them outside India. Users should not submit sensitive or confidential data unless the provider notice and their organisation permit it.

5. Failures

Provider outage, timeout, queue delay, callback interruption or output defect does not automatically create a cash-refund right. Credit reconciliation follows the Credits Policy. Creative Lab will avoid duplicate provider submission where the original outcome may still complete.

6. Output restrictions

Users are responsible for model-specific commercial-use, attribution, watermark, geographic, age, content and redistribution rules. Creative Lab may technically enforce these restrictions.

7. No endorsement

Integration does not imply that a provider endorses Maveristic, Creative Lab, Oryyn or any output. Provider names and marks belong to their owners.