xAI has just released Grok 4.1, claiming the top spot in real-world conversational AI and making it instantly available to every user for free across grok.com, X, and the mobile apps.
Introducing Grok 4.1, a frontier model that sets a new standard for conversational intelligence, emotional understanding, and real-world helpfulness.
— xAI (@xai) November 17, 2025
Grok 4.1 is available for free on https://t.co/AnXpIEOPEb, https://t.co/53pltyq3a4 and our mobile apps.https://t.co/Cdmv5CqSrb
Released on November 17, 2025, after a two-week silent rollout where it already surpassed the previous production model in 65% of blind user preference tests, Grok 4.1 now dominates the LMArena Text Leaderboard. Its thinking mode (quasarflux) scores 1483 Elo, and its fast non-reasoning mode (tensor) scores 1465 Elo, marking a 31-point lead over any non-xAI model and a significant leap from Grok 4's former #33 ranking.
The model achieves record scores of 1586 on EQ-Bench3 for emotional intelligence and 1722 on Creative Writing v3, a 600-point improvement over xAI's prior best. It also reduces factual hallucinations to one-third of the previous rate, making it the least error-prone Grok yet.
🚨Text Leaderboard Update@xAI’s Grok 4.1 (thinking) and Grok 4.1 have scaled new heights in the most competitive Text Arena:
— lmarena.ai (@arena) November 17, 2025
🔹Grok 4.1 (thinking) lands at #1 with a score of 1483
🔹Grok 4.1 follows at #2 with a score of 1465
On the Arena Expert leaderboard:
🔸Grok 4.1… https://t.co/qi6jBJ4WdT pic.twitter.com/S8nsd5X8DH
Users switching to Grok 4.1 today report noticeably faster responses and sharper quality, with more empathetic and creatively compelling outputs across emotional roleplays, writing tasks, and everyday queries.
Grok 4.1 has higher emotional intelligence, empathy, and interpersonal skills, scoring 1586 on EQ-Bench. pic.twitter.com/2V8yRZaN4O
— xAI (@xai) November 17, 2025
xAI built Grok 4.1 on the same massive reinforcement learning system used for Grok 4 but shifted focus to personality coherence, nuanced intent detection, and autonomous reward modeling at frontier scale. This delivers what the company calls the new standard in blind human preference evaluations.
Every user worldwide, including those with completely free accounts, can select Grok 4.1 right now in the model picker or let Auto mode deploy it by default. This continues xAI's pattern of shipping the world's most capable models to the broadest possible audience without paywalls.
