# Inception – A new frontier in LLM speed

> Markdown mirror of DialtoneApp's public top-site detail page for `inceptionlabs.ai`.

URL: https://dialtoneapp.com/top-sites/inceptionlabs.ai/index.md
Canonical HTML: https://dialtoneapp.com/top-sites/inceptionlabs.ai

## Summary

- Domain: `inceptionlabs.ai`
- Website: https://inceptionlabs.ai
- Description: ai readable | score 20 | purchase read only
- Label: ai_readable
- Payment surface: Not available
- Purchase boundary: read_only
- Control boundary: unknown
- Rank: 140239

## robots

~~~text
User-agent: *
Allow: /

Sitemap: https://www.inceptionlabs.ai/sitemap.xml
~~~

## llms

~~~text
# Inception

> Inception is an AI research and product company building diffusion-based large language models (dLLMs). Unlike traditional autoregressive LLMs that generate one token at a time, Inception's Mercury models generate tokens in parallel using a coarse-to-fine diffusion process, delivering 5x faster inference with best-in-class quality at a fraction of the cost. Mercury models are OpenAI API-compatible and run on standard GPUs.

Key facts about Inception and Mercury:

- Inception builds diffusion large language models (dLLMs), a fundamentally different architecture from autoregressive LLMs like GPT or Claude
- Mercury models generate tokens in parallel rather than sequentially, enabling dramatically faster inference without sacrificing quality
- Mercury is production-grade and deployed at Fortune 500 companies. It is available through AWS Bedrock, Azure Foundry, and Inception's own API
- Mercury models are OpenAI API-compatible, making them a drop-in replacement for existing LLM workflows
- Three core use cases: AI agents, real-time voice/search, and coding (autocomplete, tab suggestions, chat)
- Pricing: $0.25 per 1M input tokens, $0.75 per 1M output tokens
- The founding team includes leading researchers from Stanford, UCLA, and Cornell who pioneered foundational AI technologies including diffusion models, Flash Attention, and Direct Preference Optimization (DPO)

## Models

- [Mercury 2](https://www.inceptionlabs.ai/models): The fastest reasoning LLM and the first reasoning dLLM. Ideal for complex applications where both performance and speed are critical.
- [Mercury Edit](https://www.inceptionlabs.ai/models): A small, coding-focused dLLM optimized for code editing and extremely latency-sensitive components of coding workflows. Integrated with the Zed code editor.

## Getting Started

- [API Platform](https://platform.inceptionlabs.ai/): Sign up and get API access to Mercury models
- [Mercury Chat](https://chat.inceptionlabs.ai/): Try Mercury 2 in the browser
- [API Documentation](https://docs.inceptionlabs.ai/get-started/get-started): Quickstart guide and full API reference
- [Integrations](https://docs.inceptionlabs.ai/resources): Available integrations and deployment options

## Company

- [About Inception](https://www.inceptionlabs.ai/about): Company mission, team, and founding story
- [Research](https://www.inceptionlabs.ai/research): Published research from the Inception team
- [Blog](https://www.inceptionlabs.ai/blog): Product announcements and technical deep dives
- [Introducing Mercury 2](https://www.inceptionlabs.ai/blog/introducing-mercury-2): Mercury 2 launch announcement
- [Careers](https://jobs.gem.com/inception): Open roles at Inception

## Enterprise

- [Enterprise Solutions](https://www.inceptionlabs.ai/enterprise): Fine-tuning, private deployments, custom SLAs, and 99.5%+ uptime
- [Contact Sales](https://www.inceptionlabs.ai/enterprise#contact-sales): Get in touch for enterprise pricing and deployment options
- [Customer Stories](https://www.inceptionlabs.ai/enterprise#customer-stories): How teams are using Mercury in production

## Research Papers

- [Diffusion Models (Ermon et al.)](https://arxiv.org/abs/2010.02502): The foundational approach for modern image and video generation, co-developed by Inception CEO Stefano Ermon
- [Flash Attention](https://arxiv.org/abs/2205.14135): A key algorithm for efficient GPU utilization in LLM training and inference
- [Direct Preference Optimization (DPO)](https://arxiv.org/abs/2305.18290): One of the core approaches for aligning LLMs with human feedback
- [Masked Diffusion (MDLM)](https://arxiv.org/abs/2406.07524): Masked diffusion language models
- [d1 Reasoning](https://arxiv.org/abs/2504.12216): Reasoning capabilities for diffusion language models
- [Block Diffusion](https://arxiv.org/abs/2503.09573): Block-level diffusion for efficient text generation
- [Discrete Diffusion Guidance](https://arxiv.org/abs/2412.10193): Guidance methods for discrete diffusion models

## Contact

- [Sales](mailto:sales@inceptionlabs.ai): Enterprise and sales inquiries
- [General Inquiries](mailto:hello@inceptionlabs.ai): General questions
- [Discord](https://discord.com/invite/5VySp6ctXB): Developer community
- [X / Twitter](https://x.com/_inception_ai): @_inception_ai
- [LinkedIn](https://www.linkedin.com/company/inception-labs-ai/): Company updates

## Optional

- [Terms of Service](https://www.inceptionlabs.ai/docs/terms-of-use): Legal terms
- [Privacy Policy](https://www.inceptionlabs.ai/docs/privacy-policy): Privacy policy
- [Pricing](https://www.inceptionlabs.ai/models#pricing): Detailed model pricing
~~~

## llms-full

Not found.