The company claims it secured a top score of 89.4 in the Arena-Hard benchmark, which compares AI models based on how they respond to human prompts.
pull down to refresh
pull down to refresh
The company claims it secured a top score of 89.4 in the Arena-Hard benchmark, which compares AI models based on how they respond to human prompts.
Cool, but Deepseek’s claim to fame isn’t that it is the best AI but rather they found a way to create it far cheaper than what the rest of big tech has been doing.