AI providers · Anthropic · updated September 2, 2026

Run your AI employees on Claude

Bring your own Anthropic API key and Coordinator's AI team members run on Claude — reading vendor PDFs natively, working inside real roles and fences, billed by Anthropic directly to you at zero markup.

Free. No card. About two minutes to set up.

What you can delegate

A Claude-powered employee, not a chatbot

An agent here holds roles exactly like a person does, and its ceiling is its roles' ceiling: whatever someone with those roles could do in Coordinator, you can hand over. Claude decides how well the work gets done — not how far it reaches.

Money & books
  • Read a forwarded vendor PDF into a draft bill, validated before it touches the books
  • Watch receivables aging and draft the chase
  • Answer P&L and client-profitability questions from the live ledger
Projects & work
  • Create, file and update activities across your pipelines and boards
  • Work a queue — claim the next item, do it, hand off
  • Run the check-in round-up and keep stages honest
Time & billing
  • Log time as it works
  • Sweep unbilled hours into a draft invoice
  • Keep the time ledger clean enough to bill from
Records & forms
  • Triage form submissions into records and the right pipeline
  • Keep client records current as work happens
  • Find anything — records, contacts, history — on request
Scheduling
  • Prep today's meetings from the calendar
  • Tend booking-link intake and meeting polls
  • Keep resources and rooms straight
Chat & threads
  • Draft replies, follow-ups and summaries inside your team's threads
  • Announce what needs announcing
  • Turn a meeting thread into the follow-up tasks
The short list it can never do

Hold the Owner role. Message outside your organization before you deliberately climb the outbound ladder. Appear human in the audit trail — agent work is labeled agent work, always. Keep going after you hit pause. Everything else opens closed and is yours to grant, role by role.

The full AI-employees story — roles, fences, the outbound ladder
Why Claude

What Claude itself brings

Any provider on this site can power the delegation above. These are the reasons to pick Claude in particular.

  • Reads the file you were actually sentClaude takes a multi-page PDF as a document — no rasterising, no page-by-page screenshots. Vendor invoices are almost always PDFs, which is why Claude is our default engine for bill and receipt extraction.
  • The agent-work reputation is earnedSonnet 5's million-token context and agent tuning — at a $2/$10 price Anthropic made permanent — is what the strongest developer sentiment of 2026 keeps pointing at. Haiku 4.5 covers the high-volume end for a fifth of that.
  • One vendor, both directionsYour key powers agents inside Coordinator — and Coordinator's MCP server lets Claude (claude.ai, Claude Code, Claude Desktop) drive your workspace from the outside. Same account, both directions.
  • Models

    Claude models your agents can use

    Prices and lineup verified September 2, 2026. In the product, the picker is fetched live from Anthropic on your own key — it never goes stale, and a custom model id always works if it validates.

    Claude Haiku 4.5default

    $1/$5 per million tokens, 200K context — extraction is structured reading, and Haiku is accurate at it and the cheapest thing Anthropic sells

    Claude Sonnet 5balanced

    $2/$10, million-token context — the launch price Anthropic made permanent; the agent-workflow workhorse

    Claude Opus 5most capable

    $5/$25, million-token context — for the messiest documents and hardest calls

    Claude Fable 5.1frontier

    $10/$50 — Anthropic's frontier tier, reachable through the live picker if your key can call it (check the retention terms on frontier models first)

    Every current Claudelive list

    the picker asks Anthropic's Models API what your key can reach today — new Claudes appear without waiting on us

    Media through Coordinator's pipeline: PDFs and images. The authority on current rates is Anthropic's pricing page.

    What developers say · September 2, 2026

    Claude's API, as its users describe it

    Recurring themes from public developer forums and reviews — the likes and the complaints — and how each plays out when the provider runs inside your own workspace.

    What they like
    • Opus 5's July 2026 launch — near-frontier capability at half the old flagship price — drew one of the year's biggest positive developer receptions.
    • Sonnet 5's million-token context and adaptive thinking get credited for real agent-workflow fit, not benchmark theatre; Haiku 4.5 at $1/$5 per million tokens is the routinely named sweet spot for high-volume work.
    • The Rate Limits API (April 2026) fixed a long-standing gripe — your code can now ask what its actual limits are instead of scraping the console.
    What they complain about — and how it plays here

    Capacity is the recurring 2026 complaint: platform-wide overload (529) spikes reliably follow big model launches, with several incidents this year.

    Here: An agent here isn't married to one vendor. Keep a second provider connected and switching the brain mid-incident is a settings change — the roles, fences and workflows don't move.

    Cost-tracking surprises made the rounds — viral threads about API spend far outrunning expectations on agent workloads.

    Here: Every agent turn bills your own Anthropic key at Anthropic's prices, with usage attributed per agent in the audit trail — the spend is visible where the work happened, and Anthropic workspaces let you cap it at the key.

    The June 2026 requirement of 30-day data retention on Anthropic's newest Fable/Mythos tier drew loud backlash; Anthropic announced a rollback on September 1, 2026.

    Here: The models this page lists — Haiku 4.5, Sonnet 5, Opus 5 — were never in that covered tier and keep the not-retained-by-default posture. If you route agents to a covered model, read Anthropic's retention page first; policy here is moving week to week.

    Ours to get right

    How Coordinator treats your Anthropic key

    The section above is Anthropic's report card. This one is ours — the plumbing decisions that make handing an agent your API key boring instead of brave.

    Stored encrypted, shown as last-four

    Your key is encrypted like every credential in Coordinator, and after connecting, nothing ever displays more than its last four characters.

    Proven before it's stored

    Connecting is a live round-trip to Anthropic — a key that doesn't actually work can't be saved, so failures happen in settings, not mid-task.

    Billed by Anthropic, never by us

    Zero markup, no metering, no shared key anywhere in the system. Every agent turn bills your own account at the vendor's own prices.

    The picker asks Anthropic, not a list

    Model choices come from Anthropic's Models API on your key, so new Claudes appear without waiting on us — a baked-in list exists only as an outage floor, and a custom model id always works.

    Getting started

    From no key to a working agent in a few minutes

    Create a key in the Anthropic Console under Settings → API keys, on a workspace with credits.

    First, at console.anthropic.com
    1. 1

      Sign in at console.anthropic.com (a Claude.ai subscription is separate — the API is billed on its own).

    2. 2

      Open Settings → API keys and choose Create key.

    3. 3

      Pick or create a workspace for it. A dedicated workspace lets you set a spend limit and revoke Coordinator's access without touching anything else.

    4. 4

      Copy the key the moment it appears — the console never shows it again.

    5. 5

      Under Settings → Billing, add credits. A brand-new organisation has a zero balance and every call fails until you do.

    Then, in Coordinator — the actual screens
    Connect Anthropic (Claude)
    API key
    sk-ant-········9tk3
    Validate & save

    Key verified — a live call to Anthropic confirmed it before anything was stored. Only the last four characters are kept visible.

    Settings → AI: pick Anthropic (Claude), paste the key. A live call to Anthropic verifies it before anything is stored.

    Default model
    Claude Haiku 4.5
    Claude Haiku 4.5default
    Claude Sonnet 5balanced
    Claude Opus 5most capable
    Claude Fable 5.1frontier

    Fetched live from your key — never a stale list. Custom model ids work too.

    Choose a default model from your key's live catalogue — or set one per agent later.

    What it costs

    Billed per token on your own account. Claude Haiku 4.5 is the cheapest model that reads documents well; a one-page bill typically costs well under one US cent. Check the pricing page before committing — rates move.

    Watch out

    The commonest failure is not a bad key: it's an unfunded workspace. If the connect test says the key was rejected but you're sure it was copied correctly, check the credit balance first.

    Your key, your spend

    Every agent turn bills your Anthropic account at Anthropic's prices. Coordinator adds no markup and never holds a shared key — connect-time validation is a live round-trip against your own credential.

    FAQ

    Claude + Coordinator, asked directly

    Which Claude models can my agents use?

    Any chat model your Anthropic key can reach — Haiku 4.5, Sonnet 5 and Opus 5 are offered by default, the live picker lists what your key sees, and a custom model id always saves if it validates.

    Do I pay Coordinator for Claude usage?

    No. Coordinator is free. Your Anthropic key is billed by Anthropic directly, at their published prices, with zero platform markup.

    Can Claude read my invoices and receipts?

    Yes — natively. Claude accepts multi-page PDFs as documents, so bill extraction works on the file your vendor actually sent, not a degraded screenshot of it.

    Is this the same as Claude's MCP support?

    They're complementary. An Anthropic key makes Claude power agents inside Coordinator; our MCP server lets your Claude apps operate Coordinator from outside. You can use either or both.

    Not sure Claude is the one?

    Swap the brain, keep the team

    Your agents aren't built out of Claude — they're built out of roles, permissions and workflows. The model behind an agent is a setting: change it in a minute and the team, the history and the fences all stay put. Every provider we support gets the same treatment, and the same honest page:

    100% free — every feature, every seat

    Put Claude on the payroll

    Coordinator is free — the only bill is the one Anthropic already sends you.

    Free. No card. About two minutes to set up.