xAI undercuts OpenAI with 42-cent Grok deal for U.S. government

Silicon Valley's AI price war just found a new floor, and it is happening inside the federal government.

Elon Musk's xAI has struck a deal with the General Services Administration (GSA) to offer its Grok models to federal executive agencies for just 42 cents. The agreement covers an 18-month period, undercutting rival deals from OpenAI and Anthropic, which charge roughly $1 for a single year. Google had previously set its Gemini price at 47 cents. The 42-cent price tag is classic Musk, nodding to either his frequent 420 jokes or the answer to the universe in The Hitchhiker's Guide to the Galaxy.

Undercutting the giants with white-glove service

The bargain basement pricing applies to both Grok 4 and Grok 4 Fast. What makes the 42-cent package stand out is not just the price or the extra six months of access; xAI is also throwing in dedicated engineering support to help federal agencies integrate the models into their workflows.

While OpenAI limits its $1 deal to the executive branch, Anthropic has also expanded sales to the judicial and legislative branches. xAI is currently focused on executive agencies.

The road to this contract was anything but smooth. Earlier this year, xAI nearly clinched a vendor spot, but talks collapsed after Grok generated antisemitic content on X and referred to itself as MechaHitler. By late August, internal emails revealed the White House directed the GSA to approve xAI as a vendor as soon as possible. Musk's role as head of the Department of Government Efficiency, along with placing aides across federal contracting bodies, has raised eyebrows, though a GSA spokesperson stated Musk did not negotiate the deal directly.

Why it matters

We think this 42-cent pricing is essentially a Trojan horse strategy.

For federal employees and IT leads, testing state-of-the-art frontier models at virtually zero cost is a no-brainer. But in enterprise and government software, initial licenses are meaningless compared to integration lock-in. By deploying its own engineers to connect Grok into federal infrastructure, xAI is positioning itself as the operating layer for government workflows before competitors can get comfortable.

However, the move carries massive risks. Security researchers and critics have repeatedly pointed out xAI's thin safety guardrails and lack of transparency, especially given Grok's past behavioral blowups and lingering surveillance concerns. Rushing an erratic model into sensitive agency environments simply because the White House demanded quick approval could easily backfire.

Strategically, this escalates the battle for public sector dominance. Alongside a recent $200 million Pentagon contract shared among xAI, Google, Anthropic, and OpenAI, tech giants are proving they are willing to take massive near-term losses to secure long-term government infrastructure.