Math.inc

Math.inc is an academic research tool. Math, Inc. is building Gauss, an autoformalization system aimed at machine-generated mathematical proofs that can be formally verified. The site currently offers only an early access request form, so capability, pricing and platform details are not yet public.

Math, Inc. is a research venture working on what it calls verified superintelligence through autoformalization. The public site is currently a single early access page: visitors learn that the team is building a product named Gauss and can request a place in the queue ahead of a wider release. The company's stated direction is machine reasoning that produces proofs which can be checked, rather than plausible-looking answers that have to be taken on trust. The practical workflow starts with a short form. A visitor supplies a name, an email address and a free text answer to the question of what they want to prove with Gauss. That last field is the interesting one. It signals that access is being handed out selectively, likely to mathematicians, formal methods researchers, verification engineers and teams with a concrete theorem, lemma or specification they want machine assistance on. There is no self-serve signup, no dashboard and no published documentation beyond the request page. Autoformalization, the concept the site is organized around, means converting informal mathematics (a paper, a proof sketch, a natural language specification) into a formal language that a proof assistant can verify. That places Math, Inc. in the same broad territory as automated theorem proving tools and formal verification platforms, but with an explicitly research-first posture. The emphasis on verification is the differentiator worth watching: the output is meant to be checkable, not merely convincing. Because the product is pre-release, buyers should treat this listing as a signal to monitor rather than a tool to adopt today. No pricing, platform support, integration list or feature documentation is published. Anyone with a proof obligation that cannot wait should stay with established proof assistants and their existing libraries, and revisit Gauss once early access opens and the team publishes concrete capability details.