21–30 of 50 posts
This is exciting. Is there a free version of DeepSeek R1 that's completely US based, so we're not sending data to China? I guess you can use this to deploy it, but I'm asking for an application that would be safer to use if you're concerned about Chinese influence.
You can run it yourself: https://workos.com/blog/how-to-run-deepseek-r1-locally
This is bizarre to see the entire AI hype cycle speedrun all over on just DeepSeek. I'm trying to square the excitement over DeepSeek with its good -but not dominant- performance in evals.
IMO more anti hype for openAI who might be dominant, but are they $3500 per task (O3 high) dominant, or $200 per month dominant.
This is bizarre to see the entire AI hype cycle speedrun all over on just DeepSeek. I'm trying to square the excitement over DeepSeek with its good -but not dominant- performance in evals.
For me the excitement is that around the o3 announcement I had a feeling like we were heading to an OpenAI / Sam Altman controlled dystopia. This resets that - you can run the model yourself, you can modify it yourself, it's essentially on par with the best public models, and it gives hope that the smaller players have a fighting chance going forward. They also published their innovations bringing back some of the fe…
This is bizarre to see the entire AI hype cycle speedrun all over on just DeepSeek. I'm trying to square the excitement over DeepSeek with its good -but not dominant- performance in evals.
Also R1 (and its distilled models) expose their CoT & web interface has a websearch option too.
With the 14b distilled models, I found multiple math-related prompts where it gives the right answers almost immediately but then wastes 10 minutes making self-verification mistakes (e.g. "Write Python3 code that computes the modular inverse of a mod 2^32")
Earlier quoted context omitted.
Distilled R1 models != R1
You can run the full R1 (671B variant) locally as well so long as you have the hardware for it. `ollama run deepseek-r1:671b` will do that
Earlier quoted context omitted.
OpenAI o1 and Deepseek r1 have similar performance (o1 is a bit better at reasoning though you can see r1’s though process which you could argue trumps the competition). OpenAI o1 api cost: $60/million output tokens. Deepseek r1 api cost: $2.19/million output tokens.
So basically I can get a ~lifetime supply of r1 tokens for… $20?
r1 api can spit out 63 tokens per second
~143 billion lifetime tokens.
~$313 million for a lifetime supply of tokens.
Earlier quoted context omitted.
IMO more anti hype for openAI who might be dominant, but are they $3500 per task (O3 high) dominant, or $200 per month dominant.
Right, but Google already had models that were as good at much lower cost.
Earlier quoted context omitted.
So basically I can get a ~lifetime supply of r1 tokens for… $20?
~lifespan = 2.27 billion seconds r1 api can spit out 63 tokens per second ~143 billion lifetime tokens. ~$313 million for a lifetime supply of tokens.