Ranking of AI assistants
Claude has the highest Celsius score, 74 out of 100, ahead of Gemini (71) and ChatGPT (62).
Celsius score of the 10 AI assistants, out of 100
The 3 category scores, equally weighted. Usage and prices are published alongside and do not count towards the score. Sensitivity to weights
- 1
ClaudeAnthropic · United States74987747
- 2
GeminiGoogle · United States71944673
- 3
ChatGPTOpenAI · United States62935240
- 4
CopilotMicrosoft · United States58635853
- 5
Meta AIMeta · United States48773137
- 6
VibeMistral AI · France4074172
- 7
QwenAlibaba · China39493433
- 8
PerplexityPerplexity · United States36283942
- 9
DeepSeekDeepSeek · China35431845
- 10
GrokSpaceX Super Intelligence · United States26491713
Performance, Ethics, Environment: the detail is in the Table tab.
Scores by Projet Celsius; the sources for each criterion are on the Methodology page. Projet Celsius, AI Ranking, scores as of 11 November 2026 (methodology 1.1). CC BY 4.0 licence.

"Rankings" series on the Éclairages blog
All articles on AI →
Blog article ↗environmentWhich AIs pollute the most?10
Blog article ↗ethicsWhich AI is the most ethical?1
Blog article ↗Celsius scoreWhat are the best AIs of 2026?1
Blog article ↗useWhich AIs are the most used?1Pilotage IA, the tool that measures AI use in an organisation
Choosing an assistant for your teams involves costs, data and a footprint. Pilotage IA measures the actual usage in your organisation, shows what weighs most and quantifies what each decision changes.
For a fictional company of 480 people, as calculated by the tool.

Continue in the ranking
The full ranking →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







