Now Reading: SpaceXAI launches Grok 4.6 as cheaper rival to OpenAI and Anthropic
-
01
SpaceXAI launches Grok 4.6 as cheaper rival to OpenAI and Anthropic
SpaceXAI launched Grok 4.6 on August 12, claiming its model matches leading systems from OpenAI and Anthropic at a lower cost. The company says the model excels at agentic tasks, research, coding,…
The Story in Brief
SpaceXAI launched Grok 4.6 on August 12, claiming its model matches leading systems from OpenAI and Anthropic at a lower cost. The company says the model excels at agentic tasks, research, coding, software development and analysis, by working through many steps. Artificial Analysis places Grok 4.6 close to OpenAI’s GPT-5.6 Sol on its Intelligence Index, though early tests do not show it consistently beating top rivals.
Pricing is a key differentiator: about $2 per million input tokens and $6 per million output tokens, far cheaper than many frontier models. The model was trained longer using synthetic data and refined with reinforcement learning. It is available through Cursor and Grok Build. Early users report strong performance on long coding tasks, though some say it still falls short in certain areas.
The Indian Opinion
The launch feeds a familiar narrative: a cheaper model supposedly matching the leaders. But early benchmarks from Artificial Analysis show Grok 4.6 close to OpenAI’s GPT-5.6 Sol, not ahead. Some users on Cursor say it still trails the best in certain tasks. The real test is not price per token but whether independent evaluations confirm Grok 4.6 can sustain complex, multi-step work at scale. Until then, cheaper is not enough.
Source: newindianexpress.com
This story was synthesised by AI from the source linked above.
