Grok 4.5 is available in Grok Build, in Cursor, and from the SpaceXAI console, but not in the EU; it costs $2 per 1M input tokens and $6 per 1M output tokens
Today we are releasing Grok 4.5 together with SpaceXAI …Michael Nuñez /VentureBeat:SpaceX's Grok 4.5 launches at half the price of rivals — here's why that could rattle Anthropic and OpenAIVyom Ramani /Digit:Grok 4.5 and Cursor were trained together: It could change who codes with whatComputerworld:SpaceXAI launches Grok 4.5, touts lower coding-task costs than AI rivalsShimul Sood /Android Authority:Grok 4.5 is here to justify your X subscriptionKim Jeong-Uk /Seoul Economic Daily:xAI Launches Gr
Context & Ripple Effects
This release extends xAI’s progression from the Grok 3 API, whose listed prices were $3 per million input tokens and $15 per million output tokens, to a newer model with lower published token rates. It also follows Grok 4’s positioning as a high-performing model and the later introduction of a premium Grok 4 Heavy tier.
The notable distribution change is not just an API launch: Grok 4.5 is being offered through Grok Build, the SpaceXAI console, and Cursor, which the coverage describes as a joint training partner. Availability remains geographically uneven because the EU is excluded.
First-order effects
- Developers using Grok Build, Cursor, or the SpaceXAI console gain access to Grok 4.5 at listed rates of $2 per million input tokens and $6 per million output tokens, below the earlier listed Grok 3 rates in the coverage.
- Cursor becomes an immediate distribution point for a model trained alongside it, while EU-based users cannot access the launch through the announced channels.
Second-order effects
- The lower published price creates a sharper cost benchmark for providers competing for coding and long-running-task workloads, including Anthropic, particularly where customers compare token costs across tools.
- Embedding the model in Cursor shifts competition toward integrated coding environments: model providers need distribution inside developer workflows, while coding-tool partners can differentiate through closer model integration.
Third-order effects
- If lower-priced, workflow-specific model launches persist, frontier-model competition may increasingly hinge on the combined price, distribution, and task integration of a model rather than standalone benchmark positioning.
- The EU exclusion highlights how regional availability can fragment AI product rollouts, leaving global developer platforms to manage uneven model access across their user bases.
The trend: This is one data point in the move from selling general-purpose models through APIs toward competitively priced models distributed directly inside specialized developer workflows.