X CEO Elon Musk has placed Grok 4.6 at the top of a comparison of leading artificial intelligence models, saying it offers the strongest overall balance of intelligence, speed and cost. Musk made the claim after sharing figures attributed to Artificial Analysis that compared Grok 4.6 with GPT-5.6 H Sol (max) and Claude Opus 5 (max).
Taking to X on Wednesday, Musk posted the comparison showing Grok 4.6 (high) alongside the two rival models and declared: “Grok 4.6 is objectively #1 when considering intelligence, speed & cost.”
The figures in the comparison ranged from $0.18 (Sh23.22) to $2.34 (Sh301.86), although Musk did not explain which individual figure belonged to each model or how the final ranking was reached.
Other figures displayed in the comparison included $0.36 (Sh46.44), $0.64 (Sh82.56), $1.23 (Sh158.67), $0.41 (Sh52.89), $0.32 (Sh41.28), $0.21 (Sh27.09), $0.48 (Sh61.92), $0.87 (Sh112.23), $0.84 (Sh108.36), $0.29 (Sh37.41) and $0.40 (Sh51.60).
The figures appeared alongside the different model names and configurations, with Grok 4.6 listed as high, while GPT-5.6 H Sol and Claude Opus 5 were shown in their max configurations.
Musk's post did not give a detailed breakdown of the figures or show how intelligence, speed and cost were weighted to produce the overall ranking. The post instead presented the comparison as evidence that Grok 4.6 currently offers the strongest combination of the three measures.
Artificial Analysis was named as the source of the comparison shared by Musk. The figures show a wide gap in the costs presented, with the highest amount of $2.34 being more than 13 times the lowest figure of $0.18.
The comparison comes as developers and technology companies continue to compete over increasingly capable AI models while also seeking to keep the cost of using them within reach of consumers, businesses and developers.
By putting intelligence, speed and cost at the centre of his assessment, Musk's claim focuses on the balance between what an AI model can deliver and how much users have to pay to access it.
For businesses and other users handling large amounts of AI-generated work, the comparison also draws attention to the cost of running AI systems alongside their performance. A model that can provide strong results while using fewer resources could be more attractive to organisations managing high volumes of AI tasks.
Musk's post therefore adds to the growing contest among advanced AI models, with Grok 4.6 being presented alongside GPT-5.6 H Sol and Claude Opus 5 as competitors in a market where capability, speed and affordability are becoming key measures of value.