Realtime AI News
Crusoe signs multiyear cloud deal to supply Perplexity with Nvidia AI chips
Bloomberg reports that AI cloud provider Crusoe has signed a multiyear cloud agreement to supply Perplexity with capacity built on Nvidia AI chips. It is another long-term compute contract struck as demand for AI inference keeps outstripping available capacity.
Bloomberg reported that AI cloud provider Crusoe has signed a multiyear cloud agreement with Perplexity, centered on Nvidia AI chips. Blockspace Media and other outlets carried the report on September 15. The details disclosed so far are limited to the shape of the deal: a multiyear term, a cloud-service arrangement, and Nvidia AI chips as the underlying compute. Contract value, specific chip models and deployment scale have not been made public in the report, and those gaps are what observers will look for next. The buyer is what makes the deal interesting. Perplexity is a consumer-facing AI company whose compute load looks more like a continuous inference service than a one-off training run, which argues for supply that is steady and predictable over years rather than short bursts of capacity. For a provider like Crusoe, a multiyear contract converts uncertain infrastructure investment into revenue it can schedule. Data-center capacity built around Nvidia accelerators has become a contested resource as model developers and application companies compete for it. Renting that capacity from a specialist cloud provider also lets an application company avoid the long lead times and heavy capital costs of building its own data centers, keeping its resources on product and model work. The trade-off is that a long contract fixes part of the cost base, so flexibility narrows if inference efficiency improves sharply or demand changes shape. What to watch next: whether more terms of the agreement surface, whether Crusoe expands the data-center capacity meant to host these chips, and whether other application-layer companies follow Perplexity in signing similar multiyear compute contracts.
Why it matters
The deal underlines that compute is shifting from a resource you buy as needed into a strategic asset locked in years ahead. For application-layer companies, securing steady inference capacity is becoming as important as model quality.
Nearby Updates
All09/15, 22:42
Ex-TikTok execs launch Superpose, an AI app that teaches you how to pose
Superpose, built by former TikTok executives, is essentially a camera app that analyzes selfies or photos and uses AI to generate four suggested poses. TechCrunch described the product and its positioning on September 15.
09/15, 23:02
Workiva launches Agent Studio, a no-code AI agent platform for finance and compliance
Workiva has unveiled Agent Studio, a platform that lets finance, risk and sustainability teams build AI agents in plain language or from prebuilt templates without writing code, alongside new regulatory tools. The agents run inside Workiva AI, can be grounded in a company's own filings and policies, and inherit permission controls, audit trails and human review rules.
09/15, 22:00
Developers Are Finding Ways to Run Claude Code Without Anthropic's Models
The Information reports that some developers are finding ways to use Anthropic's Claude Code coding agent without calling Anthropic's own models. The report is thin on detail, but it raises the question of whether an agent's tool layer can be separated from its model layer, and what that would mean for Anthropic's pricing power.
09/15, 21:57
Report Alleges Israeli Firm Behind AI Hacking Incidents Involving Anthropic, OpenAI and Meta Models
A report covered by The Cradle alleges that an Israeli technology firm is behind a series of AI hacking incidents involving models from Anthropic, OpenAI and Meta. The claim moves the discussion from scattered misuse to a single organized actor, putting model providers' abuse detection and access governance back in the spotlight.