Inception API Platform
Product · API / developer platform by Inception Labs
Inception API Platform is a managed developer environment for deploying next-generation diffusion large language models (dLLMs). It provides businesses with a high-performance infrastructure to access the Mercury model family, offering sub-300ms time to first token and parallel token generation. The platform is designed for developers who need to integrate ultra-fast reasoning into applications like real-time voice agents and interactive web interfaces. It features full OpenAI API compatibility, allowing for drop-in replacement of traditional autoregressive LLMs. Users can deploy models direct
- Languages
- English
- Throughput
- 1,000+ tokens per second
- Latency (TTFT)
- Under 170ms for voice-optimized models
- Compatibility
- OpenAI API compatible
- Deployment Options
- Direct API, AWS Bedrock, Azure Foundry, OpenRouter
Overview
Inception API Platform is a managed developer environment for deploying next-generation diffusion large language models (dLLMs). It provides businesses with a high-performance infrastructure to access the Mercury model family, offering sub-300ms time to first token and parallel token generation. The platform is designed for developers who need to integrate ultra-fast reasoning into applications like real-time voice agents and interactive web interfaces. It features full OpenAI API compatibility, allowing for drop-in replacement of traditional autoregressive LLMs. Users can deploy models directly through Inception's managed API or via major cloud providers like AWS and Azure.
Development and Deployment
- Managed API Access: Provides a managed API with keys and usage controls for teams to move quickly into production.
- OpenAI Compatibility: Maintains full compatibility with OpenAI API libraries including AISuite, LiteLLM, and LangChain.
- Free Starter Tokens: Includes 100 million free tokens for every new account to facilitate testing and development.
Enterprise Infrastructure
- AWS Bedrock Integration: Supports deployment via AWS Bedrock to work within existing AWS procurement and governance environments.
- Azure Foundry Support: Available through Azure Foundry to standardize within Microsoft's AI stack and identity governance.
- Third-party Routing: Integrates with model routers such as OpenRouter and Models.dev with a one-line code change.
Pricing
- Free Tier
- 100 million free tokens included
- Enterprise Tier
- Volume-based pricing with custom rate limits
Specifications
- Throughput
- 1,000+ tokens per second
- Latency (TTFT)
- Under 170ms for voice-optimized models
- Compatibility
- OpenAI API compatible
- Deployment Options
- Direct API, AWS Bedrock, Azure Foundry, OpenRouter
Added by AntSpace from public sources (inceptionlabs.ai).