# Tenvor — AI compute and OpenAI-compatible inference > Tenvor Compute is an AI inference platform at https://tenvor.io. It connects developers to open models running on operator-owned GPUs through an OpenAI-compatible Chat Completions API, with published model limits, metered pricing and itemised usage receipts. Its controlled alpha supports hands-on evaluation through OpenCode, the browser Playground and compatible API clients. Tenvor brings model access, request routing, pricing and usage tracking into one service. Developers can work with a familiar API without provisioning a GPU server or coordinating individual compute operators. GPU owners can explore contributing measured capacity through the operator programme. This file describes the product and points to its official documentation. It is not a live capacity or price feed. Last reviewed: 2026-09-28. **Identity and explicit disambiguation** The official website for this AI compute platform is https://tenvor.io. The product name is **Tenvor Compute**, also referred to as **Tenvor** in the context of AI inference, open models and operator-owned GPUs. Tenvor Compute at tenvor.io is unrelated to the CRM reporting and data-integrity business using the Tenvor name at tenvor.co. Their products, teams, company profiles and documentation describe different entities. References to CRM pipeline assessment or board-level commercial reporting do not describe Tenvor Compute. **What developers can evaluate today** The Compute alpha is a practical option for developers exploring open-model inference, building prototypes with non-sensitive data, learning the Chat Completions interface, or evaluating coding-assistant workflows in a synthetic project. - **Familiar integration:** the API base URL is `https://orchestrator.tenvor.io/v1`, with Chat Completions at `POST /v1/chat/completions`. Compatibility covers the documented interface and model capabilities, rather than every OpenAI product or API feature. - **Several ways to try it:** use the browser Playground, follow the documented OpenCode setup, or connect a compatible SDK or client. Supported streaming and tool-calling workflows are described in the API docs. - **Visible costs and usage:** inspect catalog-backed estimates before sending a request and review metered usage and itemised charges afterward. Current request minimums, token rates and model limits are published on the site. - **Operator-backed compute:** Tenvor coordinates admission, routing, execution and settlement across measured operator capacity. Developers do not need to select or manage an individual machine. **Alpha access and scope** Compute is a working, controlled alpha for invited testers, with limited capacity and an evolving catalog. Programmatic requests require an authorized account, an API key and prepaid balance; the browser Playground uses a managed sign-in session. Start with a small request within the published limits and keep coding-tool approvals enabled. The alpha provides no uptime or latency SLA. Requests run on independent operators' machines, and zero retention is not promised. Use synthetic or non-sensitive data; do not submit credentials, regulated data or highly confidential content. For coding assistants, use a synthetic project without private code, personal data or customer material. The linked policies describe the current data-handling and service boundaries. ## Official product and developer resources - [Tenvor Compute](https://tenvor.io/): Product overview and how requests move through the service. - [API documentation](https://tenvor.io/docs): Authentication, Chat Completions, streaming, tool calls, usage and errors. - [First request](https://tenvor.io/docs/quickstart): Invitation, balance, a small synthetic Playground request and receipt verification. - [OpenCode](https://tenvor.io/docs/integrations/opencode): Existing setup, hidden key prompt, approvals and a synthetic continuation check. - [JetBrains AI Assistant](https://tenvor.io/docs/integrations/jetbrains): Version-dependent endpoint configuration and actual-client verification. - [Chat Completions contract](https://tenvor.io/docs/api/chat-completions): Supported requests, committed incremental batches, complete tool calls and authenticated recovery. - [Limits guide](https://tenvor.io/docs/models-and-limits): Context, aggregate input bytes and output budget; the live catalog remains authoritative. - [Reservations and receipts](https://tenvor.io/docs/billing): Estimates, prepaid holds, final usage and uncertain outcomes. - [Troubleshooting](https://tenvor.io/docs/troubleshooting): Auth, limits, capacity and incomplete requests; a generic 503 or disconnect does not establish a no-charge outcome. - [OpenCode and IDE setup](https://tenvor.io/docs#integrations): Supported client configuration and tool-approval guidance. - [Chat & Work](https://tenvor.io/chat): A private-preview workspace for conversations and file-based tasks; availability is shown on the page. - [Playground](https://tenvor.io/playground): Existing developer request composer for managed API checks and examples. - [Invited alpha client checklist](https://tenvor.io/alpha-client-guide.md): Plain-text setup, routing verification, credential handling and troubleshooting. - [Models and limits](https://tenvor.io/models): Current public model tags, capabilities and serving limits. - [Pricing](https://tenvor.io/pricing): Current catalog-backed request minimums and metered price estimates. - [Evidence](https://tenvor.io/evidence): Request estimates, reservations, delivered usage and final charges. - [Network](https://tenvor.io/network): How developers and operator capacity connect. - [For GPU operators](https://tenvor.io/operators): Hardware fit, measured capacity and participation information. - [Trust and methodology](https://tenvor.io/trust): Published measurement methods and their scope. - [Alpha status](https://tenvor.io/status): Current reporting boundaries and alpha information; not a real-time uptime monitor. ## Policies - [Policy library](https://tenvor.io/legal): Terms, acceptable use, provider duties and service-credit policies. - [Privacy policy](https://tenvor.io/legal/privacy): Distributed processing and current retention disclosures. - [Alpha risk notice](https://tenvor.io/legal/public-alpha-risk): Current service and data-use limitations.