.png){kind=logo}
Hey all, would love to be able to use Deepseek's LLMs. On par performa...
Hey all, would love to be able to use Deepseek's LLMs in Clay. On par performance-wise with OAI/Claude and so much cheaper. Do you have any plans to integrate their model?
Bruno R.··Great feedback! I've already surfaced this to the team and will make sure to follow up on the suggestion.
Mark L.··Wes, we've added new reasoning models in the "Content creation" category, including DeepSeek R1 and OpenAi's o3-mini. https://downloads.intercomcdn.com/i/o/w28k1kwz/1365387570/b14e941accd5214aab2595bff8ac/20250204-ai-models.png?expires=1738706400&signature=ca81be4336bfe3cb553407c69197237478a2837c532b77a9726dde66ff22c816&req=dSMhE8p2moRYWfMW1HO4zcxLPLNKnc0Zlcujzd6iTOHy0qD%2F8k1puXtqSMIX%0Ak8nE%0A Can't wait to see what you do with these!
Wes H.··Amazing, thank you!!
- Wes H.··
Any plans to integrate Deepseek's non reasoning model? (R3)
- Mark L.··
I don't know of specific plans for that one – I think the reasoning transparency was the main driver for these last updates – but hope this indicates our interest. :) Any particular use cases you have in mind for R3? Curious if it's primarily a cost/efficiency impact (as you alluded a couple week back), or if you think there are other workflows or activities that R3 specifically would unlock for you...?
- Wes H.··
Ah, thanks for clarifying. Yes, it's purely a cost/efficiency play
- Mark L.··
Makes sense! I did a quick check internally, and seems like we've done some concurrent platform work to help expand our LLM footprint more quickly in the future. So, higher potential for Llama, DeepSeek V3, etc. in the coming weeks. Didn't see any timeline details, but your vote counts! Thanks again for taking the time to nudge. I think seeing user excitement match our own helped push this forward!
