Perplexity or DeepSeek
Celsius score: Perplexity 36, DeepSeek 35 out of 100, 1 point apart; the method does not treat a gap of less than 3 points as significant. Perplexity is ahead on ethics, DeepSeek on performance and environment.
Gaps by category
- Performance28430out of 100100
DeepSeek is ahead of Perplexity on performance: 43 against 28 out of 100.
- Ethics39180out of 100100
Perplexity is ahead of DeepSeek on ethics: 39 against 18 out of 100. Largest gap: "Generated content", Perplexity 58 against 33.
- Environment42450out of 100100
DeepSeek is ahead of Perplexity on environment: 45 against 42 out of 100. Largest gap: "Low-carbon electricity", Perplexity 33 against 17.
- Use0.9 %3.4 %0% of AI sites' web traffic60
DeepSeek has 3.8 times more web traffic than Perplexity: 3.4% against 0.9% of visits to AI sites.
- Price2.50 $3.96 $0$ per million output tokens, API50
The Perplexity API costs 1.6 times less for output: $2.50 against $3.96 per million tokens. Subscriptions: €22 (Perplexity) and free (DeepSeek).
Continue in the ranking
The full ranking →AssistantPerplexity36out of 100, 8th out of 10The Perplexity entry: scores, plans, models, data.The entry →AssistantDeepSeek35out of 100, 9th out of 10The DeepSeek entry: scores, plans, models, data.The entry →Comparison2 AI assistantsPerplexity or Vibe36 against 40Celsius score out of 100Another comparison, with an assistant close in the ranking.The comparison →The hub newsletter
This month's ranking, news on the impact of AI and one figure from the AI Impact Database. One newsletter a month at most, free of charge. Past issues
