xAI's Grok 4.6 Finishes Initial Training This Week at 2 Trillion Parameters
Elon Musk confirmed Grok 4.6 is completing its initial training run, scaling up from Grok 4.5's 1.5 trillion parameters while aiming to keep the same speed and token efficiency. He positioned it as a direct answer to Moonshot AI's Kimi K3.
Elon Musk confirmed this week that xAI's next flagship model, Grok 4.6, is finishing its initial training run, scaling up from Grok 4.5's 1.5 trillion parameters to 2 trillion.
- Grok 4.6: 2 trillion parameters, up from Grok 4.5's 1.5 trillion
- Initial training expected to complete the week of July 20, 2026
- No official launch date yet, but a release as soon as August 2026 is plausible
- Musk says token efficiency and inference speed will stay close to the 1.5T model, not scale up with parameter count
- Framed explicitly against Moonshot AI's Kimi K3 (2.8T parameters), which Musk says costs roughly three times more per task
Better than our 1.5T in every way - Elon Musk, on Grok 4.6
xAI has not published benchmark numbers or a technical report, so how Grok 4.6 actually stacks up against Kimi K3 and other frontier models on coding and reasoning tasks remains unverified until it ships.
More AI news in Polish at nowosci.ai