Grok 4.5 is going public. Elon Musk confirmed on July 8 that SpaceXAI will open the model to all users starting tomorrow . The announcement follows a private beta period at SpaceX and Tesla, where the model was used internally for engineering and manufacturing workflows .
Musk described Grok 4.5 as an “Opus-class” model — a direct comparison to Anthropic‘s flagship Claude Opus 4.8 . He also claimed the model is faster, more token-efficient, and cheaper . The model is built on a 1.5-trillion-parameter V9 foundation, with supplemental training data from Cursor — the AI coding assistant SpaceX agreed to acquire earlier this year .
The timing is notable. GPT-5.6 is already rolling out to some ChatGPT users ahead of its full public launch this week . Google’s Gemini 3.5 Pro is expected on July 17 . The AI model market is about to get crowded. Grok 4.5 is joining the party just as the door opens.
The SpaceXAI Advantage
Grok 4.5 isn‘t just another model release. It’s the first major product launch under the SpaceXAI brand, following xAI‘s rebranding and consolidation into SpaceX . The model’s private beta was conducted inside SpaceX and Tesla, using real-world engineering data from rocket design, autonomous driving, and manufacturing . That‘s a different training environment than OpenAI or Anthropic can offer.
Grok 4.5’s V9 foundation is roughly three times larger than the 0.5T v8-small version currently handling Grok‘s production traffic on X . The addition of Cursor data suggests a focus on coding and reasoning tasks — the same battlefield where Claude Code and GPT-5.6 are competing .
What’s at Stake
Musk is positioning Grok 4.5 as a direct challenger to the current performance leaders: Anthropic‘s Opus 4.8 and OpenAI’s GPT-5.6. The claimed cost advantage is the key lever. If Grok 4.5 is cheaper to run than Opus or GPT, it could capture developers looking for alternatives — especially after Fable 5‘s government restrictions and GPT-5.6’s staggered release . But the model‘s performance on independent benchmarks hasn’t been released yet. Musk‘s “Opus-class” claim is currently a claim, not a verified result. The company’s long-term goal remains orbital AI compute, but Grok 4.5 is proof of concept on the ground.
P.S. Grok 4.5 is launching tomorrow — but it‘s been running inside SpaceX and Tesla for weeks. The engineers building rockets and autonomous cars have already tested it. Now the rest of us get to find out if it’s actually Opus-class, or just Musk-class confidence. Either way, the AI week just got more interesting.
