What is Gabber?
Gabber offers a visual app flow builder (App Flow Builder) that supports drag-and-drop node connections, enabling real-time fusion of multiple input streams such as video, screen, audio, and text. With a built-in sub-second orchestration engine, it seamlessly integrates inference, logical decision-making, and state management, allowing deployment from prototype to production in just a few minutes.
The platform currently covers over ten real-world use cases, including personal fitness coaching, sports coaching, AI companions, disaster early warning systems, personal stylists, screen monitors, AI research assistants, NPC creation, threat detection, interview preparation, speech feedback, poker assistance, sous-chef support, translation, and Twitch streaming assistance. Some of these features are currently locked and require additional authorization to enable.
Key features
Supports multi-stream input fusion: It can simultaneously connect video, screen sharing, microphone audio, and text inputs, enabling the creation of real-time AI applications with multimodal perception capabilities.
Provides a composable, scalable, and interoperable node system (Composable Blocks), allowing developers to freely combine modules such as visual understanding, speech-to-text (STT), text-to-speech (TTS), logical control, and state management, thereby flexibly designing any real-time AI application architecture.
Use cases and audience
AI developers, creators of multimodal applications, enterprise technology teams, AI researchers, and professionals in fitness and educational technology
- Real-time AI Personal Coach: Monitors user movements and posture, counts repetitions, and motivates users to push beyond their limits
- Real-time Multimodal AI Application Development: Integrates video, screen, audio, and text inputs to build cross-device, multi-participant real-time AI applications
- Professional-Scenario AI Assistants: Such as sports coaches, interview coaches, presentation feedback tools, poker aids, kitchen assistants, translators, Twitch streamer helpers, and more
Video streams, screen sharing, audio streams, and text inputs
Real-time visual understanding results, synthesized speech output, natural language responses, motion feedback, and behavioral recommendations
Frequently asked questions
What counts toward usage?
Billing is based on AI session hours: basic sessions $1/hour, STT speech-to-text $0.15/hour.
Is there a limit on concurrent sessions?
There is no limit on concurrent sessions.
How to use
- 01
Use the drag-and-drop App Flow Builder to visually construct applications, or reuse existing templates for a quick start
- 02
Connect visual language model (VLM), large language model (LLM), speech-to-text (STT), and text-to-speech (TTS) nodes
- 03
Combine multi-stream inputs (video, screen, audio, text) with logic and state modules to create real-time AI applications that respond in seconds
- 04
Deploy to the Gabber cloud environment or run locally—get up and running in just minutes
Benefits and limitations
Benefits
- Ultra-low latency
- Exceptional expressiveness and lifelike sound
- Lower cost compared to competitors
- Multimodal building blocks
- Technology-stack-agnostic integration
- High-definition cloning with unlimited length and consistent quality
- Emotionally expressive speech with emotion tags
Limitations
- Language-specific support may require a separate request
- HD Clone requires direct contact with them.
Pricing and availability
Hourly billing applies: basic sessions $1/hour; STT $0.15/hour; TTS can be self-hosted or provided via Gabber (charged at cost); LLM costs vary depending on the model; orchestration services are free; there is no limit on concurrent sessions; only community support is offered. The Enterprise edition supports custom integrations, dedicated infrastructure, volume discounts, and priority support.
- Pricing source
- https://gabber.dev/
- Pricing last updated
- 2026-10-30
Prices and limits may change. Check the official website for current terms.
Official website & sources
Updated on 2026-10-30
