Musk tweets on September 24: SpaceXAI to ship Fable/GPT-6-class model in 2-3 months
Musk replied to critics on X saying that if acceleration holds, SpaceXAI will reach 'pole position' in about 6 months, with hardware roll-out speed as the core barrier.
On September 24, Elon Musk posted on X in response to criticism that SpaceXAI was falling behind in the AI race, saying: 'I am cautiously optimistic that SpaceX will have a Fable/GPT-6 level model in 2 to 3 months.' The remark came three days after SpaceXAI shipped Grok 4.7, which is priced at $2 input and $6 output per 1M tokens and is positioned for coding and long-horizon knowledge work, with independent and company-reported scores mixed across software and electrical engineering benchmarks.
When user Alex commented 'it's over for SpaceXAI,' Musk responded in three parts. First, SpaceX's AI effort is only 3 years old versus Anthropic's 6 and OpenAI's 10, and if acceleration holds the company will reach 'pole position' in roughly 6 months. Second, for most production workloads, 'you don't need (and it would be cruel to put) Newton-level intelligence in your toaster' - marginal intelligence above what a task requires adds little. Third, hardware is the core constraint: bringing massive compute online rapidly is 'incredibly difficult,' and SpaceX has demonstrated unusual capability here and will continue improving.
SpaceXAI's public narrative leans on 'acceleration' and 'hardware roll-out speed' as its moat. Musk said model roadmap cadence remains monthly, with next-generation foundation models to be trained from scratch targeting Anthropic Fable and OpenAI GPT-6-class intelligence.