Skip to content

The datasheet for every AI agent

Vecbase

Verified

Summary

Vecbase is a hosted workspace where each AI agent is set up like an employee: its own instructions, model, sandboxed cloud computer, file-space grants and connected accounts. Work arrives as chat tasks and lands as real files in a shared library.

Best for
Small teams or solo operators who want pre-built AI agent roles, shared files, and OAuth-driven workflows under one credit-based subscription.
Not for
Larger organizations that need fixed monthly pricing rather than credit-based billing, or teams that want broad independent editorial coverage before committing.
Free tier
PartlyPricing publishes only Plus (2,000 credits/month) and Pro (5,000 credits/month), both paid. Every /tools/* page's model menu badges Vecbase 1.0 Lite as Free and Vecbase 1.0 Pro as Pro, so a free tier appears to exist in-product, but its allowance is never stated.
Visit Vecbase CompareFind alternatives

What it does

Writes code
YesAgents build web pages and other code from a written brief and hand back editable source files.Explain the offer, audience, and action. The Vecbase Agent shapes the message, builds a responsive page, and gives you a live preview with source…vecbase.com/tools/landing-page
Debugs and fixes
YesTakes a stack trace or error log and returns a root-cause report with a proposed patch and verification checklist.Paste a stack trace or add an error log. The Vecbase Agent follows the evidence, tests likely causes, and returns a root-cause report with a patch…vecbase.com/tools/bug-investigator
Writes tests
YesA dedicated unit-test task writes tests against the uploaded code and runs them where the environment allows.The Agent prepares meaningful unit tests that fit the codebase and keeps the evidence and open questions clear for review.vecbase.com/tools/unit-test-generator
Reviews code
YesReviews a PR link or diff and returns findings with file and line references, severity and suggested patches.Paste a PR link or diff. The Vecbase Agent reads the surrounding code, runs relevant checks, prioritizes real risks, and proposes testable changes.vecbase.com/tools/code-review
Whole-codebase work
PartlyA repository link or nearby files can be supplied as context; the vendor claims no whole-repo indexing and no repo-wide migration.For code review, add a pull request, repository link, or code context, explain what matters, and let the Vecbase Agent prepare a review with evidence…vecbase.com/tools/code-review
Ships it
PartlyDelivers a live preview plus deployable source files; nothing on vecbase.com claims it deploys to your own hosting or merges to your repo.Vecbase designs a responsive, accessible landing page and returns a live preview with deployable source files.vecbase.com/features/tools
Done-for-you service
NoSelf-serve software you operate yourself; no build engagement, contract or managed service is offered.Sign in and grant the capabilities you need — minutes to integrate, no contracts to sign, no APIs to wire up.vecbase.com/features/services

How much it does unattended

Runs autonomously
YesScheduled agents run unattended and push results to the file library, Slack or email; two consecutive failures auto-pause the schedule.Daily 9 AM reports, Monday briefs, six-hourly checks — agents run on schedule and drop results in the file library or your inbox.vecbase.com/features/schedules
Multi-agent
YesAgents delegate sub-tasks to each other inside one workspace; the Plus plan lists multi-agent workflows as included.An agent can delegate sub-tasks to another agent — Writer pulls the report from Researcher to draft a blog, Operator asks Coder to run a data script.vecbase.com/features/agents
Agent permissions
YesEach agent is limited to the tools, file spaces and connections you grant it, and sensitive actions can be held for human approval.An Agent can only use the tools, files, and connected accounts available to it. You decide which of these resources each Agent may use.docs.vecbase.com/introduction

Which models it runs on

Claude
YesAnthropic models are in the hosted catalogue and Claude Opus is the default shown for the research agent.Top-tier models from Anthropic, OpenAI, Google, plus regional leaders Kimi, DeepSeek, and our own Vecbase.vecbase.com/features/models
GPT
YesOpenAI models are in the same catalogue; GPT is the model shown for the coding agent.Researcher runs on Claude Opus for reasoning, Coder leans on GPT for code, Writer goes with Kimi for prose — each agent picks independently.vecbase.com/features/models
Open models
PartlyKimi and DeepSeek sit in the hosted catalogue, but the vendor never describes any model as open-weight and publishes no way to supply or run your own.Top-tier models from Anthropic, OpenAI, Google, plus regional leaders Kimi, DeepSeek, and our own Vecbase.vecbase.com/features/models
Model choice
YesEach agent is assigned its own model and can be switched mid-conversation without losing the chat.Researcher runs on Claude Opus for reasoning, Coder leans on GPT for code, Writer goes with Kimi for prose — each agent picks independently.vecbase.com/features/models

Where you use it

In your editor
NoPublished surfaces are the web app and a macOS desktop client; no IDE or editor extension exists on the site or in the docs.Run the same AI team and workspace you use on the web in a focused desktop window, with secure system integration and automatic updates.vecbase.com/download
On the command line
NoThe desktop client is the same web product in a window; no terminal client is offered; bash exists only as a tool inside the agent's sandbox.The desktop app uses the same Vecbase product, account, workspace, agents, and visual language as the web.vecbase.com/download
In your pipeline
PartlyAn agent can act on GitHub through a scoped OAuth grant, including opening pull requests; no CI runner or pipeline integration is published.Each agent gets its own grant, e.g. Researcher reads email only, Coder opens GitHub PRs.vecbase.com/features/connectors
In a browser
YesThe hosted web workspace is the primary surface; a signed macOS desktop build runs the same product, account and agents.Run the same AI team and workspace you use on the web in a focused desktop window, with secure system integration and automatic updates.vecbase.com/download

Whose machine it runs on

Self-hosted
NoThe only two work environments are a Temporary workspace and a Dedicated cloud computer, both run for you by Vecbase; no on-prem or self-hosted option is published.Vecbase offers two simple choices.docs.vecbase.com/agents/work-environmen…

Not established: Open source

What it costs to run

How it meters
YesMetered in credits drawn down by model tokens, sandbox compute, storage and each managed capability; plans bundle 2,000 or 5,000 credits a month.Everything your agents consume: AI model tokens, compute time, storage, speech-to-text, text-to-speech, image generation, document parsing, and deep…vecbase.com/pricing
Free tier
PartlySign-up and checkout are self-serve with no sales call, but both published plans are paid credit subscriptions.Choose Plus at 2,000 credits/month with 2,000 usage credits, or Pro at 5,000 credits/month with 5,000 credits and persistent cloud computers.vecbase.com/pricing
Bring your own key
NoVecbase supplies and meters the model and capability calls itself and bills them as credits; no bring-your-own-key or own-cloud path is published.All capability calls aggregate into a single monthly bill — no need to manage each vendor's billing.vecbase.com/features/services

Not established: API access, MCP server

Buying it for a team

A company can buy it
YesBought at organization level: one subscription and balance covers every workspace, cancellable from organization billing.One subscription and credit balance cover every workspace in your organization. Credits reset each month and do not roll over.vecbase.com/pricing
Seat model
NoPriced per organization on a monthly credit allowance; no seat count, per-seat price, floor or ceiling appears on the pricing page.One subscription and credit balance cover every workspace in your organization. Credits reset each month and do not roll over.vecbase.com/pricing
Pooled budget
YesA single credit balance is shared across every workspace, with usage broken down per agent and per capability.One subscription and credit balance cover every workspace in your organization. Credits reset each month and do not roll over.vecbase.com/pricing
Admin controls
YesConnector scopes, capabilities and file spaces are granted per agent and revocable, and sensitive actions can require human approval.When granting Gmail to an agent, choose read-only, send-only, or both. Revoke any time. Every call is logged and audit-friendly.vecbase.com/features/connectors
Audit log
PartlyAgent tool calls are logged and every scheduled run leaves a record; no administrative audit log of member, permission or billing changes is documented.Every call is logged and audit-friendly.vecbase.com/features/connectors

Not established: Single sign-on

What happens to your code

Getting out
PartlyUnused credits expire at the end of each month; generated files stay in the File Library, which offers downloads and S3-compatible access.Credits reset each month and do not roll over.vecbase.com/pricing

Not established: Opt out of training, Data residency, Certifications

28 sourced claims on this page. Checked 2026-09-03. How we check.

Alternatives

Where we would send a reader instead, in that listing's own words.

  • Botpress

    AI agent platform with a visual studio, a code-first TypeScript ADK and a helpdesk, priced per conversation.

  • Dify

    Open-source platform for building agentic workflows, AI agents and RAG knowledge pipelines.

  • Lindy

    AI teammate in Slack, email and meetings, sold per seat with a shared credit pool.

  • MindStudio

    No-code visual builder for AI agent workflows, published as web apps, APIs, webhooks or MCP servers.

  • Relevance AI

    Hosted low-code platform for building, evaluating and running teams of business AI agents.

  • Sim

    Open-source AI workspace for building, deploying and running agents across 1,000+ integrations.

  • n8n

    Workflow automation platform for building and running AI agents on a visual canvas, with code where needed.

Filed under

AutonomousClaudeGPTMulti-agentOpen models

Sources and updates

Advertise here

Reach buyers mid-decision. Reach builders choosing their next agent. Promote your brand with a display placement or bring your listing into focus with Featured.

Explore owner options →Advertise on this page →

The digestFree

Which agents actually ship.

What we re-checked, what got added, and one number from the index. Tuesdays.

110 agents trackedOne-click unsubscribe