Add DeepSeek-R1 and DeepSeek-R1-Zero Models to Merlin for Advanced Capabilities
in progress
Samuel Jackson
It would be amazing to see the DeepSeek-R1 and DeepSeek-R1-Zero models integrated into Merlin! These models, available on Hugging Face, are cutting-edge and could significantly enhance Merlin's capabilities for deep reasoning and zero-shot tasks.
Adding them would provide more flexibility for advanced workflows, making Merlin an even more powerful tool for research, problem-solving, and creative tasks. Any plans to integrate these models in the future?
endu
Merged in a post:
DeepSeek R1 - Open Source Reasoning Model
Reda Izo
They was a previous closed post here from November 2024, with closed reason being IRR what does that even mean? Anywho Deepsek jsut released the model hopefully we can get it a wayy cheaper credit unit than O1 thank you
endu
Hi Reda,
Thanks for bringing this up! 😊 Yes, we’re excited to share that we’re adding the DeepSeek-R1 model to Merlin soon. Stay tuned for updates—we're thrilled to make this happen! 🚀
Appreciate your enthusiasm! 🙌
Reda Izo
endu Nice! are we getting 671B or distill models?
endu
in progress
endu
Hi Samuel,
Thanks so much for sharing this! 😊 The DeepSeek-R1 and DeepSeek-R1-Zero models sound fantastic, and we agree that they could bring incredible value to Merlin. We're excited to let you know that we’ll be adding them! 🎉
Appreciate your suggestions—keep them coming! 🚀
Samuel Jackson
endu Let's go! You guys are doing a really great job. Thank you!
Samuel Jackson
API Access & Pricing
Use DeepSeek-R1 by setting model=deepseek-reasoner
$0.14 / million input tokens (cache hit)
$0.55 / million input tokens (cache miss)
$2.19 / million output tokens