Elon Musk Says Grok 4.7 Lands in 10 Days and Will Beat Every Model
Elon Musk mentioned Grok 4.7 can be launched to the general public in 10 days and that the mannequin will surpass each AI mannequin presently out there.
The launch follows Grok 4.6, which SpaceXAI shipped on August 12. It additionally comes as OpenAI introduced that its new Astra mannequin can be launched quickly.
Follow us on X to get the most recent information because it occurs
SpaceX Data Sits on the Center of Grok 4.7
Musk has constructed the case throughout a run of posts. He said in mid-August that preliminary coaching had completed and that SpaceX firm information was being incorporated by way of supplemental coaching.
Musk had earlier detailed the structure. Grok 4.7 runs on a 2.1 trillion parameter base, up from the 1.5 trillion parameter basis behind Grok 4.6. He mentioned the bigger mannequin runs barely extra slowly whereas utilizing tokens extra effectively.
He additionally named the rival he expects to path him.
“Grok 4.7 will exceed all present fashions. That mentioned, Anthropic is a good firm and will in all probability launch improved fashions quickly. However, the SpaceX coaching corpus is so superior & distinctive that I might be shocked if any mannequin is best at real-world engineering than 4.7,” the (*10*) learn.
Meanwhile, the discharge cadence has tightened. SpaceXAI took Grok 4.5 public in July and shipped Grok 4.6 on August 12.
Grok 4.6 Benchmarks Set the Bar
Grok 4.6 was constructed on its predecessor, Grok 4.5. The firm mentioned the mannequin has a “specific deal with long-running brokers and extra bold interactive and visible work.”
According to figures shared, the mannequin scored 61 on the AA Intelligence Index, stage with GPT-5.6 Sol Max and behind Claude Fable 5 Max at 62. Grok 4.5 scored 56.
Grok 4.6 led GDPVal-AA v2 with 1753. Yet it managed 26% on Terminal-Bench v3.0, properly behind GPT-5.6 Sol Max at 34.6%.
Grok 4.5 set a similar pattern. It topped Artificial Analysis’s AutomationBench-AA at 51.4% whereas costing $0.34 per job. However, it logged 0.63 guardrail violations per job, above Claude Opus 4.8’s 0.55.
Musk has now connected a agency quantity to the discharge. Whether SpaceXAI ships unbiased evaluations will decide how far the engineering declare travels.
Subscribe to our YouTube channel to observe leaders and journalists present professional insights
The put up Elon Musk Says Grok 4.7 Lands in 10 Days and Will Beat Every Model appeared first on BeInCrypto.