SAN FRANCISCO — Elon Musk says xAI's Grok 4.5 has reached the leading edge of artificial intelligence, claiming the model now sits alongside Anthropic's Claude Opus 5 on what he called the "Pareto frontier" — the set of models that no rival beats on both capability and cost at once. In the same breath, he laid out a relentless release cadence: Grok 4.6 and Grok 4.7 are already on the calendar, roughly two and four weeks away.
The claim, made on X, underscores how quickly xAI has closed the gap with the most established labs. Grok 4.5 launched earlier this month as xAI's first model tuned specifically for coding and agentic work, and independent benchmarks placed it near the top of the industry at a fraction of the price of comparable frontier systems.
What the Pareto Frontier Means
In plain terms, being on the Pareto frontier means there is no other model that is both smarter and cheaper. Plenty of systems can match Grok on one axis, but Musk's argument is that none beats it on both. For developers weighing capability against per-token cost, that combination is the whole game, and it is why Grok has been climbing adoption charts inside coding tools. It is the same efficiency edge that has let xAI push Grok into everyday products from spreadsheets to inboxes.
xAI has leaned hard into real developer data to get there, training on live coding sessions and benchmarking against tasks that measure what a model can actually accomplish inside a real codebase over a long session, not just trivia.





