1 / 1Cogalloy Pro USB AI Voice Interaction Box
A 72 × 40 × 23 mm, 37 g USB/Type-C AI voice interaction box for host-connected robot projects. It combines a noise-reducing microphone, high-fidelity speaker, CI1302 command-wake processor and driver-free USB 2.0 audio. It is a voice front end—not a standalone LLM computer or motor controller.
- USB/Type-C voice interface: Connect as a USB 2.0 audio device for recording and playback on a compatible Windows, macOS, Linux or Android host.
- CI1302 command-wake path: 2 MB Flash and matched firmware support a defined wake word and command vocabulary instead of an unspecified free-form assistant.
- Compact, mountable enclosure: 72 × 40 × 23 mm, about 37 g, with a 62.5 × 30 mm mounting-hole pattern for robot integration.
- Audio and safety are separate from robot control: The box supplies audio and recognition; host software must validate intents and retain an independent manual stop path.
Price
US$39.99
No account or payment needed. We usually confirm availability, lead time, shipping, and total price within 12 hours.
Or email us directly: sales@cogalloy.comShipping options and final delivery cost are shown at checkout.
VOICE HARDWARE
One compact box for pickup, playback and command wake-up.
The Pro enclosure combines a noise-reducing microphone, high-fidelity speaker, USB audio and CI1302 command-wake processing. It is a robust voice front end for a host-connected robot—not a generic audio claim detached from the actual robot system.
- Integrated microphone input and speaker output
- CI1302 voice processor with 2 MByte Flash
- 3.0 W per channel speaker output into 4 Ω BTL
MECHANICAL AND USB
Mount it deliberately, then use Type-C as the data and 5 V path.
The enclosure measures 72 × 40 × 23 mm and weighs about 37 g. Its 62.5 × 30 mm mounting pattern gives a clear starting point for brackets. Use Type-C USB 2.0 with a host that can provide stable 5 V and audio support.
- DC 5 V power through the Type-C port
- USB 2.0, driver-free audio on supported host systems
- Confirm screw diameter, cable bend radius and acoustic opening after assembly
FIRMWARE-FIRST
Define the wake word and command vocabulary before calling it voice control.
Wake word, language, command set and returned command data are supplied by the selected firmware. Flash through Type-C with the CI1302 target, test the serial/audio path, then validate a small vocabulary in the actual acoustic environment.
- Use the firmware that matches the target robot and language
- RST button enters firmware-update mode after the proper port is selected
- Reference range: up to 5 m quiet / about 1 m noisy; site test is required
HOST AND ROBOT SAFETY
Keep AI services and motion authority in the host application.
A Raspberry Pi, Jetson or PC interprets command data, links online ASR or a large language model when desired, and applies robot-state and safety checks. Never bind an uncertain phrase directly to motor power or movement.
- Large-model and cloud features are host/software dependent
- Map recognised commands to explicit, testable robot intents
- Keep an independent manual or emergency stop route for moving robots
Product details
Everything you need to evaluate the fit.
Clear product information, technical specifications and compatibility details—kept separate from the purchase decision.
Overview
A USB voice front end for host-connected robots
The Cogalloy Pro USB AI Voice Interaction Box is a compact voice interface for a Raspberry Pi, Jetson, PC or compatible robot controller. It combines a noise-reducing microphone, high-fidelity speaker, USB audio and CI1302-based command wake-up in one mountable enclosure. Connect it through its Type-C port as a USB 2.0 device; use the host and matched firmware to turn recognised phrases into robot actions.
This SKU is the Pro USB enclosure, not the older 48 × 24 × 10.5 mm, 4-pin I²C voice board. Do not reuse legacy-board wiring, dimensions, weight or I²C claims for this product.
What the box does — and what the host still does
The box supplies the audio input/output path and command-wake recognition. A connected host supplies the robot application, behaviour mapping and any online ASR, cloud service or large-language-model integration. It is not a standalone ChatGPT/LLM computer and it does not directly drive motors, servos or batteries.
Use explicit command intents and safety checks between speech recognition and motion. Keep an independent manual stop method whenever a robot can move.
Mechanical, electrical and audio interface
The enclosure is 72 × 40 × 23 mm and weighs about 37 g. Its mounting-hole pattern is 62.5 × 30 mm centre-to-centre; verify the supplied fasteners and clearance against the finished robot. Power is DC 5 V through Type-C. The USB 2.0 audio path is driver-free on Windows, macOS, Linux and Android according to the matched host configuration.
The onboard audio path provides microphone input and speaker output. The documented speaker output is 3.0 W per channel into 4 Ω BTL. Treat the installation location, cable routing, host USB power budget and acoustic environment as part of the integration design.
Firmware and commissioning
The voice processor is CI1302 with 2 MB Flash. Command vocabulary, wake word, language and returned command data are firmware-dependent. The matching firmware is flashed through Type-C; the standard tool flow selects the CI1302 target, selects the appropriate COM port and uses the RST button to enter update mode. Wake word, language and command vocabulary must be confirmed from the Cogalloy-matched firmware supplied for the delivered configuration.
Start with USB audio recording/playback and a wake-word test, then map a small, unambiguous vocabulary to non-motion feedback. Validate every motion command at low speed with an emergency/manual stop path available. Reference recognition distance is up to 5 m in quiet conditions and about 1 m in noisy conditions; test in the actual enclosure and room before deployment.
Technical specifications
SKU
COG-AI-VOICE-PRO
Product
Cogalloy Pro USB AI Voice Interaction Box
Power supply
DC 5 V via Type-C
Firmware path
Type-C firmware update; select CI1302 target and the correct serial port, then enter update mode with RST
Audio hardware
Integrated noise-reducing microphone and high-fidelity speaker; audio input and output
Flash capacity
2 MByte
Product weight
About 37 g
Speaker output
3.0 W per channel (4 Ω BTL)
System boundary
The connected host provides robot behaviour, online ASR/cloud/LLM integration and motion safety logic
Voice processor
CI1302
Recognition mode
Command wake-up; vocabulary, language and response data depend on the flashed firmware
Product dimensions
72 × 40 × 23 mm
Not a substitute for
A host computer, robot controller, battery/power system, motor driver, emergency stop or application software
Mounting-hole pattern
62.5 × 30 mm centre-to-centre; verify fastener diameter and clearance on the delivered unit
Host / audio interface
Type-C USB 2.0; driver-free USB audio on Windows, macOS, Linux and Android with a compatible host configuration
Reference recognition distance
≤5 m in a quiet environment; ≤1 m in a noisy environment; validate at the intended installation site
Shipping & support
Choose your destination at checkout to see the available shipping methods. Need integration advice before ordering? Our support team can help you choose compatible components.
In the box
Package contents
Cogalloy Pro USB AI voice interaction box
72 × 40 × 23 mm voice box with Type-C USB interface.
Type-C data cable
USB data and 5 V power connection for the matched host.
M4 × 8 mm mounting screws
Mounting hardware shown for the standard compatible installation kit.
M4 nuts
Matching mounting hardware shown for the standard compatible installation kit.
Host computer, robot controller, battery, motor drivers and cloud/LLM service
Not included. The exact delivered packing list and optional accessories are confirmed by the order line.
Expert support
Frequently asked questions
Is this the older 4-pin I²C Cogalloy Pro AI voice board?
No. COG-AI-VOICE-PRO is the 72 × 40 × 23 mm, about 37 g Pro enclosure with a Type-C USB 2.0 interface. The old 48 × 24 × 10.5 mm, 4-pin I²C-board data does not apply to this SKU.
Does the voice box include a standalone ChatGPT or large-language model?
No. It provides audio and command-wake capabilities. A Raspberry Pi, Jetson or PC plus the chosen application, network/API service and credentials are required for online ASR, cloud or LLM features.
How do I connect it?
Use a Type-C data cable to connect the box to a compatible host USB port. Supply DC 5 V through that path and first verify recording/playback plus the serial device on the actual host.
Can I change the wake word or commands?
Yes, when the selected CI1302 firmware supports it. Flash the matched firmware through Type-C, select the correct serial port and use the documented RST/update workflow. The wake word and vocabulary are firmware-specific.
What recognition range should I expect?
The reference value is up to 5 m in a quiet environment and about 1 m in a noisy environment. Enclosure shape, microphone location, people, reverberation and vocabulary affect results, so validate the deployment site.
Can a recognised phrase drive a robot directly?
Do not connect recognition directly to motor authority. Let the host application map the phrase to a known intent, check robot state and safety conditions, and keep a separate manual/emergency stop route.
What is included with the listed module?
The standard compatible installation set is the voice box, Type-C data cable, four M4 × 8 screws and four M4 nuts. The host, robot electronics, battery and any cloud/LLM service are excluded; the final order line remains the packing-list authority.