Elon Musk said SpaceXAI plans to release Grok 4.6 around August 7. Grok 4.7 could follow several weeks later. The update outlines two new model releases as SpaceXAI competes with OpenAI, Anthropic, and China’s Moonshot AI.
Tesla CEO revealed the plans in response to Vercel CEO Guillermo Rauch. He described Grok 4.6 as a 1.5T model with significant improvements in supervised fine-tuning and reinforcement learning.
Elon Musk Advances Grok in the Global AI Race
Elon Musk highlighted that Grok 4.7 is expected to be a 2.1T model. The new version will surpass Grok 4.6 in all respects except the serving speed. The later version will also provide increased token efficiency.
Also Read: Meta & BlackRock 2026 Texas Deal Supercharges Crypto Compute
According to Elon Musk, the August 7 timeline was just an estimate. The launch date of Grok 4.7 was not specified by him. Information about pricing and access was also omitted from the announcement.
The information came after the benchmark conducted by Rauch of Grok 4.5. According to his cybersecurity benchmark, Grok 4.5 was the most powerful solution in terms of the combination of price and performance. The conclusion was made taking into account the models and prices included in his benchmark.
As Rauch stated, Grok 4.5 was ten times cheaper than Sol by OpenAI. The cost of the model was estimated at 5.7 times lower than that of Opus 5 by Anthropic. In addition, it was 2.2 times cheaper than Kimi K3 by Moonshot AI.
In the benchmark, Grok 4.5 had almost the same performance as Kimi K3. However, Rauch rated Sol as the best model. According to Rauch, Sol was still better than Opus 5.
SpaceX AI launched Grok 4.5 on July 16. It was announced as a coding and agent-based task and knowledge work model. The price of Grok 4.5 is estimated at $2 per million input tokens and $6 per million output tokens.
Both announcements come amid increasing competition between American and Chinese AI developers. Moonshot AI launched Kimi K3 on July 16. It was announced as an open-weight model with 2.8 trillion parameters.
Why Washington Is Reviewing Open-Weight AI
According to Axios, the Trump administration is contemplating limiting some Chinese advanced AI models. Among those models singled out by the report is the Kimi K3.
Later on, Nvidia, Microsoft, Meta, among others, endorsed an open letter against wide-ranging bans on open-weight AI. They believed open models foster innovation and American superiority in the industry.
Tesla CEO Elon Musk gave the letter his full support. OpenAI CEO Sam Altman backed US leadership in both open-source and proprietary AI models. OpenAI, which was absent from the initial list, later joined as a signatory.
OpenAI and Anthropic have also sought a consistent review process of their models, which have advanced cybersecurity and national security features. According to a report, such models can be evaluated by the federal department up to 30 days before their release.
Also Read: Bitcoin Price Eyes $75K as Fed and CLARITY Act Set the Stage
