MK1 Technologies

LLMs like GPT4o often struggle with complex queries on large documents. But why? Our new benchmark spotlights a key deficiency—comprehension—that worsens as context grows. The good news: We show how our custom LLM CAMeL can augment a frontier model, vastly improving comprehension. Reach out if you have Q&A tasks that are not well served by current long context and RAG systems. Also stay tuned for more results! mk1.ai/blog/long-cont…

Read the full @tensorwave x @mkoneai post here: tensorwave.com/blog/405-billi…

We're excited to be collaborating with @tensorwave pushing the performance for LLM inference on @AMD Instinct MI300X Stay tuned, more to come!

We're very excited about this new partnership with @tensorwave! This is just the beginning! @AMD businesswire.com/news/home/2024…

Congratulations to our former colleagues @neuralink!

Watch 32 @AMD MI210 cards 🔥 on Tensorwave’s GPU cloud. Summarizing 82 articles / sec using MK1 Flywheel inference engine, hitting max TDP!!! Can’t wait for 80 MI300x cards per node. What will you run? Learn how to get the most from your GPUs at @mkoneai and @tensorwave

What’s the fastest you can run LLM inference on NVIDIA hardware? Hint: It’s with MK1 Flywheel. Get started for free on @modal_labs today!

Obsessed with optimizing performance down to the instruction? We're hiring!

Run MK1 Flywheel on @modal_labs for the fastest LLM inference 🚀 Experience faster responses and lower costs with just a few lines of code. Get started with free credits and instant GPU access. Learn more on our blog: mk1.ai/blog/modal

Launching Flywheel, MK1's enterprise LLM inference engine. Now out of closed beta and ready to make AI go brrr! You can try MK1 Flywheel now on AWS SageMaker or reach out if you're ready to deploy at scale. Learn more about our progress on our blog post: mkone.ai/blog/mk1-flywh…