--- title: "xAI's Grok 4.8 entered training 57 days late with no price card; Anthropic pays SpaceX $1.25 billion a month" description: "The 57-day wait bought a serving stack xAI wrote itself, and the delay ends with no scores and no price. For a SpaceX shareholder, the unpublished price card for 4.8 is the next receipt." publisher: "The Inference" section: "Model" published: 2026-11-10T15:57:04.726Z modified: 2026-11-10T15:57:04.726Z canonical: https://theinference.org/article/xai-s-grok-4-8-entered-training-57-days-late-with-no-price-card-anthropic-pays-spacex-1-25-billion-a-month language: en keywords: "XAI, Grok, Anthropic, SpaceX, Model Pricing, Musk" --- # xAI's Grok 4.8 entered training 57 days late with no price card; Anthropic pays SpaceX $1.25 billion a month > The 57-day wait bought a serving stack xAI wrote itself, and the delay ends with no scores and no price. For a SpaceX shareholder, the unpublished price card for 4.8 is the next receipt. [xAI](https://theinference.org/markets/companies/xai) put Grok 4.8 into reinforcement learning on November 10, 57 days after [Elon Musk](https://theinference.org/markets/people/elon-musk) said it would start RL that week, and it arrives with no benchmark scores and no price card [1]. Musk's explanation is that xAI spent the time "dumping JAX/SGLang" and rewriting its entire training and serving stack in C/C++ [1]. For a [SpaceX](https://theinference.org/markets/companies/spacex) investor, what matters is not the evals but what the model costs to serve, and what it sells for. ## The rewrite is a margin decision The delay reads as an investment in the serving layer. Back on September 14, Musk had already described Grok 4.8 as "a 2.5T model trained with our new C++ software stack" [1]; this week's post extends that rewrite from training into inference, the step that answers users. The old stack ran on JAX, a research project rather than an official Google product, and SGLang, an open-source serving framework [1]. Owning that code is the margin fix for a business that reaches [roughly 118 million monthly users and collects from only about 2.4 million paying SuperGrok subscribers](https://cryptobriefing.com/grok-nears-100-million-monthly-users/), xAI's paid subscription tier [2]. Musk admitted in early November, [as we reported](https://theinference.com/article/a-judge-called-xai-s-turbines-what-they-are-and-now-grok-s-power-bill-has-a-court-date), that serving Grok Bot costs more than those subscribers pay [3]. Price tells the same story. Grok 4.7, launched September 21, sells at $2 per million input tokens and $6 at output [1]. [Anthropic](https://theinference.org/markets/companies/anthropic) cut Claude Fable 5.2 to $8 and $40 on November 9, the eve of its own IPO; [Google](https://theinference.org/markets/companies/alphabet)'s catalog listed Gemini 4 Carbon at $4 and $20 on November 10 [1]. A lab that undercuts its rivals on price needs its own code to turn electricity into answers cheaply. *Chart: **Grok 4.7 undercuts the new frontier models on input and output tokens** Column chart comparing per-million-token input and output list prices for Grok 4.7, Gemini 4 Carbon and Claude Fable 5.2* *List price per million tokens, USD, as published: Grok 4.7 launch page (September 21, 2026), AWS price file for Claude Fable 5.2 (November 9, 2026), Google Cloud catalog for Gemini 4 Carbon (November 10, 2026), via FourWeekMBA. [1]* ## Who pays xAI today is Anthropic xAI sits inside SpaceX, which has filed to go public; its website footer already reads SpaceX Super Intelligence LLC [1]. SpaceX's prospectus says [Anthropic agreed to pay $1.25 billion a month for compute](https://fourweekmba.com/ai-grok-4-8-in-rl-after-a-57-day-wait-as-xai-dumps-jax-for-cpp/), a package that includes roughly 325,000 [Nvidia](https://theinference.org/markets/companies/nvidia) GPUs [1]. That customer is the reason the price card matters: SpaceX's AI segment absorbs most of the company's capital spending, $16.4 billion of $19.6 billion in the third quarter, and needs revenue that covers it [1]. ## The number to watch is the price 4.8 ships at If 4.8 ships on code xAI wrote itself at 4.7's prices, the AI segment can serve cheap tokens profitably and rent the same capacity to the labs it undercuts. If the stack underperforms, the 57 days and that capital spending have no product to show for them. Musk's rollout to Bot users this week will show whether the foundations hold; the price card will show whether they pay. ## Takeaway Grok 4.8 entered training 57 days late with no scores and no price, and the cost math that matters to a SpaceX shareholder now hangs on what the model ships at. ## Sources 1. [FourWeekMBA, 'Grok 4.8 in RL After a 57-Day Wait as xAI Dumps JAX for C++', 10 November 2026](https://fourweekmba.com/ai-grok-4-8-in-rl-after-a-57-day-wait-as-xai-dumps-jax-for-cpp/) 2. [Crypto Briefing, 'Grok nears 100 million monthly active users as xAI's chatbot keeps climbing', 9 November 2026](https://cryptobriefing.com/grok-nears-100-million-monthly-users/) 3. [The Inference, 'A judge called xAI's turbines what they are, and now Grok's power bill has a court date', 3 November 2026](https://theinference.com/article/a-judge-called-xai-s-turbines-what-they-are-and-now-grok-s-power-bill-has-a-court-date) --- xAI's Grok 4.8 entered training 57 days late with no price card; Anthropic pays SpaceX $1.25 billion a month — The Inference. Canonical: https://theinference.org/article/xai-s-grok-4-8-entered-training-57-days-late-with-no-price-card-anthropic-pays-spacex-1-25-billion-a-month