Elon Musk-owned xAI has begun rolling out Grok 4.1, an upgrade to the Grok 4 model. The release improves quality and speed, with free access for all users to enhance large language model performance.
xAI launched two versions of the new model: Grok 4.1 and Grok 4.1 Thinking. Both incorporate incremental improvements over prior iterations. Free users can access these models immediately, while paying customers benefit from reduced usage limits.
xAI states that Grok 4.1 reduces hallucination rates by a factor of three compared to previous models. This advancement positions the release among xAI’s strongest offerings to date. Hallucinations, where models generate inaccurate information, represent a key challenge in large language model development, and this reduction stems from targeted training enhancements.
Performance against competitors like GPT 5.1, which launched recently with gains in performance and emotional intelligence, remains unconfirmed through direct comparisons. GPT 5.1’s updates focus on refining response accuracy and contextual understanding in interactions.
LMArena’s Text Arena provides early performance data on Grok 4.1. This open-source tool enables side-by-side, blind, and randomized comparisons of large language models. Users submit prompts, and the platform pairs responses anonymously to rank models based on human preferences.
Early benchmarks from Text Arena place Grok 4.1 (thinking) at number one on the Arena Expert leaderboard with a score of 1510. Grok 4.1 ranks at number 19 with a score of 1437. These results show both models achieving new positions in the competitive Text Arena environment.
The benchmark platform observed, “This is a 40+ point improvement since Grok 4 fast, which landed in the Arena just two months prior.” Grok 4 fast entered the leaderboard in September 2025, establishing a baseline that Grok 4.1 now exceeds substantially.
🚨Text Leaderboard Update@xAI’s Grok 4.1 (thinking) and Grok 4.1 have scaled new heights in the most competitive Text Arena:
🔹Grok 4.1 (thinking) lands at #1 with a score of 1483
🔹Grok 4.1 follows at #2 with a score of 1465On the Arena Expert leaderboard:
🔸Grok 4.1… https://t.co/qi6jBJ4WdT pic.twitter.com/S8nsd5X8DH— lmarena.ai (@arena) November 17, 2025
