MatterAI

MatterAI's current product page presents an inference API alongside terminal coding, an IDE and code review. This is a set of developer services, not evidence for the older catalog label's claim of universal super intelligence. The page names OrbCode, Orbital and MatterAI Reviews as different parts of that offering.

Choose the service that fits the task

The inference section advertises an OpenAI-compatible interface for open-source and third-party models. It lists model identifiers and token-price columns, along with caching and introductory-credit promotions. A listed provider name does not independently verify that provider's affiliation or every model's current availability.
The page displays two different API base hosts in its examples. Check the current service documentation before integrating rather than copying an illustrative request unquestioningly. OpenAI compatibility is an interface claim, not proof that every upstream endpoint or parameter is supported.
Displayed latency, throughput and live counters are not measurements performed for this listing. The page labels its telemetry illustrative, and several counters appear as zero in the capture. No model request, speed comparison or price-discount calculation was tested.

Keep coding permissions under review

OrbCode is described as a standalone terminal CLI with streaming, edit and command approvals, headless operation and compatible hooks. The page labels that CLI MIT-licensed; this does not establish free or unrestricted use of the separate hosted inference service.
Orbital is presented as an IDE with multi-file generation and background agents. MatterAI Reviews describes pull-request analysis, explanations and possible fix commits. Review the requested repository permissions and proposed changes before enabling an agent to edit code, run commands or submit commits.
Automated checks can be another review input, but the site's bug-reduction, time-saving and competitor-comparison claims are not independent proof that a change is safe to merge.

Clarify storage and commercial terms

The page describes secure multi-tenant storage while elsewhere using a zero-data-retention heading to explain non-training of private code. Non-training and non-retention are different commitments. Confirm the applicable storage, retention, deployment and model-provider terms for the actual product selected.
Enterprise access-control and private-deployment options are advertised, but their availability and certification details were not audited here. Do not assume every download, API account and review subscription shares the same security or billing terms.