AI Tool

Inception Mercury Voice Review

Inception Mercury Voice is a diffusion LLM (dLLM) tuned to power voice agents, enabling natural conversation with low latency.

shipped Oct 1, 2026chatbotpaid

Domain rating71Monthly visits2.4K/mo

chatbotvideocode

Why it matters

1Diffusion LLM (dLLM) tuned for voice agents.

2Features a 128,000 context window.

3Achieves median time to first answer token under 320 ms.

4Supports up to 50,000 output tokens.

About Inception Mercury Voice

Business Model

Usage-Based (Pay Per Use)

Usage Pricing

$0.00 per million tokens per token

Platforms

API

Target Audience

Enterprise customers and developers in voice applications

Pricing Plans

Launch Special

$0.20 per million input tokens

• 50% off launch pricing
• Real-time response
• Natural conversation experience

Regular Pricing

$0.40 per million input tokens

Output Tokens Launch Special

$0.75 per million output tokens

• 50% off launch pricing
• Real-time response

Output Tokens Regular Pricing

$1.50 per million output tokens

Cost Examples

• About $0.009 per minute of conversation

Leadership

Firas Trabelsi

Yanis Miraoui

Samar Khanna

Xinyu Zhao

Gokul Gunasekaran

Emily Liu

Kenan Hasanaliyev

API Docs

Specs

API Docs

View Documentation →

API Available

Yes, public API

overview

What is Inception Mercury Voice?

Inception Mercury Voice is a diffusion LLM (dLLM) tool that enables voice agents to engage in natural conversation with low latency. It can reason, call tools, and follow long system prompts while outperforming competitors on various benchmarks.

features

Key Features of Inception Mercury Voice

Inception Mercury Voice is engineered with specific capabilities to facilitate advanced voice agent interactions. Its architecture is based on a diffusion LLM (dLLM) and is optimized for conversational AI.

Diffusion LLM (dLLM) specifically tuned for voice agents.
Enables natural conversation with low latency, with a median time to first answer token under 320 ms.
Possesses reasoning capabilities with three configurable settings: low, medium, and high.
Supports tool calling and can follow long system prompts.
Offers a 128,000 context window for extensive conversational memory.
Provides proprietary function calling for custom integrations.
Supports multimodality, processing both text and audio inputs.
Capable of generating up to 50,000 output tokens.
Outperforms competitors on various high-quality conversational benchmarks.
Offers enterprise compatibility through an OpenAI API-compatible interface.

use cases

Who Should Use Inception Mercury Voice?

Inception Mercury Voice is designed for enterprise customers and developers focused on building and deploying advanced voice applications requiring high performance and natural language understanding.

Automated Drive-Thru Ordering: Businesses seeking to automate and enhance customer experience in fast-food environments.
Financial Institution Call Handling: Financial services providers aiming to improve efficiency and accuracy in customer service calls.
AI Phone Agents for Live Customer Calls: Organizations requiring AI-powered agents to manage and respond to live customer inquiries.
Voice Agents: Developers and companies building sophisticated voice-enabled applications and interfaces.

how to use

How to Use Inception Mercury Voice

To begin using Inception Mercury Voice, developers can access its capabilities via an API. The platform provides documentation to guide integration and deployment.

1Review the API documentation available at https://www.inceptionlabs.ai/docs.
2Obtain API credentials from Inception Labs.
3Integrate the Inception Mercury Voice API into existing applications or new projects.
4Configure reasoning settings (low, medium, high) as required for specific use cases.
5Utilize supported integrations such as LiveKit, Pipecat, Vapi, or Retell for enhanced functionality.

pricing

Inception Mercury Voice Pricing & Plans

Inception Mercury Voice operates on a usage-based pricing model, with distinct rates for input and output tokens. Special launch pricing is available for a limited period.

Launch Special (Input Tokens): $0.20 per million input tokens.
Regular Pricing (Input Tokens): $0.40 per million input tokens.
Output Tokens Launch Special: $0.75 per million output tokens.
Output Tokens Regular Pricing: $1.50 per million output tokens.

Enjoying this? Get one like it in your inbox each morning.

one email a day · unsubscribe in two clicks · no third-party tracking

Pros

+Specialized diffusion LLM (dLLM) architecture optimized for voice agents.
+Achieves low latency with a median time to first answer token under 320 ms.
+Offers a large 128,000 context window for extended conversations.
+Includes proprietary function calling for custom tool integration.
+Supports multimodality, processing both text and audio inputs.
+Provides configurable reasoning settings (low, medium, high) for varied use cases.

Cons

−Pricing is usage-based, which may lead to variable costs depending on volume.
−Specific performance benchmarks against all listed competitors are not detailed.
−Requires API integration, which may necessitate developer resources.
−The 'proprietary' nature of function calling might limit customization compared to open standards.