The translation industry is evolving in another important direction: translation is increasingly becoming a software infrastructure layer.
Instead of asking users to purchase a dedicated application or translation device, developers can integrate translation directly into their own products.
OneMeta’s VerbumSDK is an example of this model.
The company launched VerbumSDK On-Prem / Edge in August 2026, allowing real-time translation and transcription systems to operate within a customer’s own infrastructure. OneMeta says organizations can train models using their own data and terminology rather than relying entirely on generic models.
Why On-Premise Translation Matters
Enterprise customers often have requirements that are different from ordinary consumers.
A company may need to protect:
- Customer conversations
- Technical terminology
- Internal documents
- Meeting content
- Industry-specific vocabulary
- Sensitive business information
For these customers, sending all communication to a public cloud may not always fit their technical or security requirements.
An on-premise or edge architecture provides another option.
According to OneMeta, VerbumSDK On-Prem / Edge is designed to run inside customer-owned infrastructure, including client-owned servers and edge devices.
Industry-Specific Translation Is an Important Opportunity
Generic translation models are designed for broad language usage.
However, different industries have different terminology.
A medical company may need medical vocabulary.
A manufacturer may need equipment terminology.
A construction company may need safety terminology.
A legal company may require highly specific terminology.
This is particularly relevant to OEM translation-device development.
A translation device designed for general consumers can focus on travel and everyday conversation.
A B2B translation device could instead be optimized around:
Manufacturing + technical terminology
Construction + safety communication
Hospitality + customer service
Healthcare + multilingual communication
International trade + business negotiation
Hardware and Translation Infrastructure Can Work Together
For a physical translation device, the AI engine can be considered one part of the overall system.
The complete product can contain:
Microphones → Speech recognition → Translation engine → AI processing → Screen/audio output
The hardware determines how the user interacts with the system.
The translation infrastructure determines how the language is processed.
This separation creates significant flexibility for OEM manufacturers.
A company can customize the hardware design, user interface and physical controls while integrating different AI or translation services according to the target market.
F10 and F80 as Dedicated Translation Platforms
Our F10 and F80 devices demonstrate the hardware side of this model.
F10 combines 139-language real-time translation with 27 offline language packages, photo translation, recording-to-text, conversation recording, bilingual subtitles and two-way simultaneous interpretation.
F80 provides more than 150 online languages and nine dialects, 27 offline languages, SIM-card connectivity, video-call translation, photo translation, text-input translation and recording translation.
The purpose is not simply to provide a translation API.
A physical device can turn those AI capabilities into a dedicated user experience designed around real-world communication.
From Translation Product to Translation Platform
The development of technologies such as VerbumSDK points toward a larger industry trend.
Translation may increasingly become an infrastructure capability that can be integrated into:
- Smartphones
- Enterprise software
- Customer-service platforms
- Meeting systems
- Translation devices
- Education platforms
- Healthcare applications
- Industrial equipment
For OEM manufacturers, this means future product development may involve not only hardware customization, but also selecting the right AI infrastructure and designing how it interacts with the physical device.
Translation is becoming a platform capability—and hardware remains one of the most important ways to turn that capability into a practical product.