Beeble Account Info

Beeble SwitchX bills you per generation in USD, and the API won't quote you a price before you hit queue. That makes this little diagnostic node oddly essential: Beeble Account Info is the one node in the pack that never spends a cent, and it's the difference between knowing you have headroom and discovering your spending limit mid-batch.

It's a read-only wrapper around two Beeble API endpoints. The cheap, fast one - GET /v1/account/info - returns your account email, billing type, spending limit, and, importantly, your rate-limit usage (RPM) and concurrent-job usage against their limits. The slower one - GET /v1/account/billing - only runs when you flip include_billing ON, and adds the current period's total spend, a per-meter usage breakdown, and the remaining budget, all in USD (Beeble doesn't expose any other currency, so don't go looking for an EUR toggle).

The inputs

There are only two, and you'll barely touch them:

• include_billing (BOOLEAN, default off) - turn it on when you want the spend breakdown. It's an extra remote call and it's rate-limited at 5 RPM, so don't wire it into a polling loop.
• trigger (STRING) - a text widget you can edit to nudge a re-fetch. The author also overrides ComfyUI's change detection so the node re-executes on every queue, meaning the numbers you see are live, not cached from an hour ago.

What comes out

This is an output node, so the human-readable result shows up in its own little panel:

• summary (STRING) - the formatted account state: email, spending limit, RPM usage / limit, concurrent jobs / limit, and (with billing on) period spend, remaining budget, and per-meter usage.
• raw_json (STRING) - the full {info, billing} payload, for when you want to feed the numbers into something else or check a field the formatter didn't surface.

Installing and using it

Same drill as the rest of the pack: install comfyui-beeble-ai via ComfyUI Manager, or git clone it into ComfyUI/custom_nodes and pip install -r requirements.txt, then restart. The node needs a valid Beeble API key - same resolution as the other nodes in the pack: the BEEBLE_API_KEY environment variable first, then a config.toml at ~/.config/comfyui-beeble-ai/config.toml (or %APPDATA%\comfyui-beeble-ai\config.toml on Windows). Getting a key at developer.beeble.ai/api-keys is the only real setup step.

Things that trip people up

Two practical gotchas worth knowing. First, the billing call can fail or return a shape the formatter doesn't recognize - the node doesn't crash, it just appends a note and dumps the raw JSON so you can read the response yourself. That's intentional, not a bug. Second, "diagnostic and free" only means no generation credits; it still needs a valid key, and repeated billing fetches count against the same 5 RPM read limit that SwitchX polling uses. If you're running Account Info alongside several SwitchX nodes on one account, you can eat the read budget surprisingly fast - 15-second polls at 5 RPM leave exactly zero room for much else, so use include_billing sparingly and turn it off once you've confirmed the budget.

Drop it at the front of a long workflow, queue once, glance at the concurrency and remaining-budget lines, and you'll know exactly whether that 100-prompt sweep is going to blow up on you - or blow a hole in your wallet.