# ModelRush

## Kurzbeschreibung
**“AI Companion API for Chat, Image, Video and…” / “One integration, intelligent routing, transparent billing.”**


ModelRush is a unified API platform for text, image, video, audio, and voice models. Developers can access multiple models and capabilities through an OpenAI-compatible endpoint, including chat completions, image generation/editing, video generation, speech, transcription, private uploads, and webhooks.

## Claim
One integration, intelligent routing, transparent billing

## Geeignet für
- API Integration
- Audio / Voice
- Automation / Workflows
- Images
- Design / Graphics
- Marketing & Advertising
- Programming / Software Development
- Texts / Content
- Video

## Kernfunktionen
- API Integration
- Image editing
- Image generation
- Embeddings
- Function calls
- AI inference
- LLM API
- Multimodal AI
- Reasoning model
- Voice output
- Language model
- Voice cloning
- Text generation
- Transcription
- Video generation
- Video editing

## Preismodell
- **free:** **Agent Basic Trial**
Limited-time trial of the Agent Basic plan; card required and automatic renewal after the trial period unless canceled. (modelrush.ai)
- **subscription:** **Agent Basic** Personal agent with a monthly runtime allowance, workspace file storage, media storage, and outgoing bandwidth allowance. 


**Agent Pro** Higher runtime, workspace, media, and bandwidth allowance for more intensive agent use. 


**Agent Max** Largest agent plan with unlimited monthly runtime and expanded file, media, and bandwidth allowance.
- **other:** **Model API Usage** Model calls are not included in the agent membership and are billed separately via prepaid credits based on actual usage. 


**Prepaid **Credits Credit packages for API usage, model calls, media generation, and agent workflows. 


**Chat Completions API** OpenAI-compatible chat interface for synchronous or streaming text responses. (modelrush.ai)


**Image / Video / Audio APIs** Image generation, image editing, video generation, text-to-speech, audio transcription, uploads, webhooks, and prediction queries. 


**Enterprise Zero Data Retention** Only for compatible routes and only with a written Enterprise agreement or DPA; Customer Content is then not intentionally persistently stored in active memory.

## DSGVO und Datenschutz
**Gesamteinschätzung:** Conditional

**On-Prem / local hosting: indirect / not available**


The website did not describe any on-premise, local, or self-hostable deployment. The documentation describes an API service operated by ModelRush; local execution on customer systems is not indicated on the website.


**Private Cloud / data center: partial**


There are references to written enterprise arrangements, region control, and zero data retention via 'written enterprise order or data-processing addendum'. However, a dedicated private cloud or isolated single-tenant environment in the EU/EEA is not specifically described on the pages found.


**EU SaaS / managed: partial**


ModelRush is clearly documented as a managed SaaS/API service. The website mentions regional execution and region endpoints, but on the pages found it does not establish general EU/EEA data residency or specifically named EU/EEA data centers for the service as a whole.


**Hybrid: indirect / not available**


A hybrid operating model with partial local/private-cloud processing and partial external processing was not described on the website.


**DPA / DPA: partial**


The website repeatedly refers to a 'data-processing agreement' and describes that ModelRush acts as a Processor/Service Provider for customer content according to customer instructions. However, a directly accessible, publicly linked full DPA text or execution process was not specifically provided on the pages found.


**No training: partial**


The Privacy Policy states that ModelRush does not use 'Customer Content' to train its own models unless there is explicit opt-in. At the same time, the website notes that third-party providers process content according to provider-specific terms and that training/logging may vary depending on model, region, plan, and route. Therefore, a generally training-free end-to-end path is not broadly established for all routes.


**Open source / transparency path: indirect / not available**


No open-source components, no openly disclosed self-hostable parts, and no explicit open-source/sovereignty path were found on the website. There is documentation, API transparency regarding models and regions, as well as export-/deletion-related information, but no substantiated open-source approach.


**Data processing**


The website describes ModelRush as a unified API gateway/routing service for its own and third-party models. Customer content may be sent to the selected model provider or the provider determined by routing. According to the Data Retention Policy, standard routes store prompts, references, and uploads in ModelRush-controlled active systems for up to 7 days as a rule; according to the website, generated outputs and provider file references are removed from the active ModelRush database no later than after 24 hours. Operational Metadata may be stored for up to 13 months. For compatible enterprise routes, zero data retention is provided via written agreement. According to the Privacy Policy, international data transfers outside the user's country are possible.


**Conclusion**


For a European directory, ModelRush is not documented as a clearly fully EU/EEA-resident standard SaaS. Positive aspects are the processor role, DPA references, deletion and retention rules, and opt-in for training its own models. Due to provider-dependent onward transfer, possible international transfers, missing publicly clearly named EU/EEA data centers, the absence of a published subprocessors list, and the lack of generally established EU data residency, from an EU/EEA perspective its use is at best **conditionally** GDPR-compliant, not uniformly and clearly safeguarded.


**Sources**


- [https://modelrush.ai/privacy](https://modelrush.ai/privacy)
- [https://modelrush.ai/terms](https://modelrush.ai/terms)
- [https://modelrush.ai/data-retention](https://modelrush.ai/data-retention)
- [https://modelrush.ai/data-deletion-policy](https://modelrush.ai/data-deletion-policy)
- [https://modelrush.ai/docs/api/regions](https://modelrush.ai/docs/api/regions)
- [https://modelrush.ai/docs/api/models](https://modelrush.ai/docs/api/models)
- [https://modelrush.ai/pricing](https://modelrush.ai/pricing)
- [https://modelrush.ai/docs](https://modelrush.ai/docs)
- [https://modelrush.ai/agents](https://modelrush.ai/agents)
- [https://modelrush.ai/integrations](https://modelrush.ai/integrations)

The website documents several privacy-relevant components for use within the EU/EEA: ModelRush describes a privacy policy, the role as a processor for customer content, data deletion, retention periods, as well as a data processing agreement in the enterprise context. At the same time, the website shows that content may be transmitted to third-party providers depending on the selected model and region, and that international transfers outside the EU/EEA may occur. A clear, generally valid proof of EU/EEA-exclusive processing or a blanket EU data residency commitment for the entire service is not stated on the website. For EU/EEA users, GDPR-compliant use therefore appears possible only *under conditions*, for example through careful model/region selection and, if applicable, a written enterprise agreement for Zero Data Retention or stronger commitments.


**Positive**


Positive aspects include the published privacy policy, the explicit processor/service provider role for customer content, references to data processing agreements, deletion and retention rules, as well as the statement that ModelRush does not use customer content to train its own models unless an explicit opt-in has occurred. In addition, the documentation refers to regions, regional availability, and Enterprise Zero Data Retention by written agreement.


**Negative**


Negative or limiting is that, according to the website, customer content is shared with selected model providers and their logging, retention, and training may vary depending on the model, region, plan, and route. The website also states that ModelRush and its service providers may process information outside the user's country and use Standard Contractual Clauses or other mechanisms for this purpose. A specific EU/EEA server location for the overall service, a published subprocessor list, and a clear general EU data residency commitment are not stated on the website.


**Server location**


No general central server/data center location for ModelRush is stated on the website. Instead, the website shows region-dependent execution depending on the model. On the pricing page, for example, execution regions such as 'us-virginia' and 'cn-beijing' are visible; the API documentation mentions region support without guaranteeing general EU/EEA data residency on the pages found. EU/EEA data centers are not specifically evidenced on the pages found.

## Hosting und Daten
- **On-Prem / lokales Hosting:** unknown
- **Private Cloud / Rechenzentrum:** teilweise / indirekt
- **EU SaaS / Managed:** teilweise / indirekt
- **Hybrid:** unknown
- **AVV / DPA:** teilweise / indirekt
- **Kein Training auf Kundendaten:** teilweise / indirekt
- **Open-Source / Transparenz-Pfad:** unknown

## Standort
**Land:** USA

**Taxonomie:** USA

Moonveil AI Inc.
2018 Sand Hill Road, Unit A
Menlo Park, California 94025
USA

## Vorteile
- One API for text, image, video, audio and voice
- OpenAI-compatible integration for chat endpoints
- Live endpoints for chat, image, video, speech/transcription, uploads and webhooks
- Regional routing/availability information for each model
- Request tracing and billing metadata
- No training on customer content by ModelRush-owned models without explicit opt-in

## Nachteile
- No publicly verifiable legal entity or company headquarters found
- No publicly available DPA found, only a reference to data-processing agreement support
- Third-party models process content under their own provider terms
- Prompts, references, and uploads are actively stored on standard routes for up to seven days; additionally, encrypted archives may exist without automatic expiration
- Sensitive data, biometric data, health, financial, or children's information should not be intentionally submitted
- Some “Spicy” models require age verification and carry increased rights/consent risks

## Quellen
- Offizielle Website: https://modelrush.ai/pricing

## Letzter Datenstand
2026-09-22

## Originalseite
https://kifox.ai/en/ki-tools/modelrush-en/
