Vectoree

Vectoree gives your AI agent a backend. Everything an agent needs to ship — models, auth, database and storage — lives in one cloud, from demo to production.

• Model Plaza — One project key across providers (xAI, OpenAI, Google, Anthropic, DeepSeek and OpenAI-compatible). Browse the catalog, probe from the CLI, and call chat completions from your server. Start with vectoree/auto.
• Auth — Email and Google sign-in for your app users, configured in the dashboard.
• Database — Postgres tables, schema and SQL from the dashboard or vectoree db; your agent can create a table without spinning up a server.
• Storage — Buckets and file uploads for avatars, assets and user content, public or private.

How it works: your coding agent installs the Vectoree skill (set up https://vectoree.ai/SKILL.md), logs in and links a project with the CLI (npx @vectoree/cli login), picks a model, then calls the OpenAI-compatible gateway (https://vectoree.ai/api/v1) or creates tables. Works with Cursor, Claude Code, Codex and other coding agents.

Pricing: pay for tokens, no monthly fee. New accounts get $5 welcome credit after CLI login + link; then top up the org wallet via Stripe, billed per model. Details: https://vectoree.ai/#pricing

Why choose Vectoree

Vectoree gives your coding agent a single backend instead of wiring provider SDKs, Clerk or Auth0, a spare RDS instance, and homemade S3 together yourself. One project key reaches many models through an OpenAI-compatible gateway. The same console covers Auth (email and Google for your app users), Postgres, and object storage. Your app keeps the OpenAI SDK and only changes base URL and key. Pricing is pay-as-you-go for tokens with no monthly fee. From demo to production, one cloud.

Company: VECTOREE INC.

FAQ

What is Vectoree? Vectoree is a backend for coding agents and the apps they ship. Model calls, user login, Postgres, and object storage sit in one console behind one project key.

Who is Vectoree for? Startup teams with several projects under one org, individual developers who want models, auth, database, and storage on one key, and people who start from a console template and let Cursor, Claude Code, or Codex wire Vectoree.

How is it different from calling OpenAI directly? One provider gives you that provider’s models only. Vectoree’s Model Plaza uses one project key across many providers and an OpenAI-compatible gateway (https://vectoree.ai/api/v1), and the same project also gets Auth, Postgres, and storage.

Does Vectoree include auth, database, and storage? Yes. Auth is email and Google sign-in. Database is Postgres. Storage is buckets for avatars, assets, and user files, public or private.

How does pricing work? No monthly fee. Sign-up basics are free; after you link a project you get $5 welcome credit. Later usage bills from the organization wallet (Stripe top-up). Exact rates: https://vectoree.ai/#pricing

Which coding agents work? Cursor, Claude Code, Codex, and other agents that can install a skill and run CLI commands.

How do I start? (1) Install the skill: set up https://vectoree.ai/SKILL.md. (2) Login and link: npx @vectoree/cli login --api-url https://vectoree.ai. (3) Pick a model (start with vectoree/auto). (4) Call the gateway with the OpenAI SDK or create tables / upload files.

Can I keep my existing code? Yes. Point the OpenAI SDK baseURL at https://vectoree.ai/api/v1 and switch the API key to your Vectoree project key.

Why not just use InsForge or a raw BaaS? InsForge and similar BaaS tools can be alternatives in the same “backend for apps” space. Vectoree is built around coding agents: one project key for a multi-provider model gateway plus Auth, Postgres, and storage the agent can bind in one flow.