Inception Labs
An AI company focused on the research and productization of diffusion-based language models, and the developer of the Mercury model series.
Timeline (last 90 days)
Inception Labs released Mercury 3, a diffusion LLM, claiming it surpasses GPT-6 Luna and Claude Haiku 5.5 on benchmarks and runs 4 to 8 times faster; generation speed reaches 1,068 tokens/second on NVIDIA GPUs, and the context window expands from 260K to 1 million tokens.
Inception Labs is the developer of Mercury.
Inception Labs' diffusion model achieves tokens/second on coding tasks far exceeding any Transformer or even hybrid architectures such as Jamba.
All Facts (6)
Inception Labs 的个人投资者包括 Andrej Karpathy 和 Eric Schmidt
85%UnverifiedInception Labs 于 2024 年正式亮相，获得了微软风投、英伟达风投、Snowflake Ventures、Databricks 等机构的投资
85%UnverifiedInception Labs 的创始团队由斯坦福、UCLA 和康奈尔大学的教授组成，曾主导过 Flash Attention、决策 Transformer 及直接偏好优化（DPO）等研究
80%UnverifiedInception Labs released Mercury 3, a diffusion LLM, claiming it surpasses GPT-6 Luna and Claude Haiku 5.5 on benchmarks and runs 4 to 8 times faster; generation speed reaches 1,068 tokens/second on NVIDIA GPUs, and the context window expands from 260K to 1 million tokens.
50%UnverifiedInception Labs is the developer of Mercury.
50%UnverifiedInception Labs' diffusion model achieves tokens/second on coding tasks far exceeding any Transformer or even hybrid architectures such as Jamba.
50%