Spend

Spend is the front of the workspace’s money: g1t.sh/<workspace>/-/spend, first under Money in the Workspace sidebar. It says what was spent over a period, where it went, the budgets that hold it, and what each task cost. Every figure on it comes from the workspace’s usage and agent records; nothing is estimated.

Usage has every meter in detail, and Billing and plans has the plan, the spend limit, payments and the statement. Spend links to both.

Whose spend

View	Who sees it	What it counts
Workspace	Owners and billing managers	Everything the workspace used, every product, and every agent’s work.
You	Everyone	What agents spent working for you: your chats with agents and the sessions you started.

Owners and billing managers switch between them with Workspace and You at the top of the page. Everyone else sees You.

The period

Choose This month, Last month, 30 days or 7 days. Days are UTC. Budgets are always this month’s, whatever the period.

The figures at the top

View	Figures
Workspace	Spent: usage at price over the period, every product, the same figure as Usage. Agents: the agent product within it, every agent’s replies, sessions and runs on repositories at price, from the same ledger. Charged: what the workspace was charged this month after included usage, credit and any discount, the same figure as Billing, against its spend limit when it has one. How spend is counted defines each.
You	Agents for you over the period. Your budget this month, and what is spent against it. Your chats with agents: the replies agents wrote you.

By day shows the period one day at a time. Hover or focus a day for its amount.

Where it went

Slice the period’s spend with the tabs:

Slice	Shows
Agents	Each agent’s share of the agent product, from the ledger: its replies, sessions, runs on repositories and sandbox time at price. @g1t is g1t’s own work on your repositories. A session’s helpers and subagents count on the agent that started it. Open one for its own Spend tab. In the You view, the agents service’s count of what agents did for you.
People	Who asked, from the ledger. Routines, agent-to-agent work and runs from before attribution have no asker. Workspace view only.
Channels	Where it was asked: each channel, direct messages together, and channels you aren’t in together, without their names.
Models	Each model’s share.
Kind of work	Chat replies, sessions, routines, helping colleagues and subagents.
Products	Every product at price, from Usage: Agent, Sandboxes, AI Gateway, Deployments, Git & storage, Packages, Security & quality and Search. Workspace view only.
Extensions	Coming: what each installed extension runs, once extensions can be installed.

A slice is part of the page’s address (?by=channel), so you can share it.

Spend less, keep quality

Every week, g1t checks each agent’s finished sessions and asks one question: would a cheaper effort have done the work as well? The answers are under Spend less, keep quality, on Spend for every agent and on each agent’s Spend tab for that agent.

How a suggestion is made:

1. What counts. Each agent’s own sessions that finished in the last 28 days, at the effort level each recorded, with what each was charged as budgets count it: the model at the provider’s price and the agent rate. A session is accepted when it finished with nobody having to step in: no one steered it, and it was not stopped or failed. Helpers and subagents count as part of the session that started them.
2. What is compared. The level the agent runs at now (for Auto, the level most of its sessions ran at) against the level below.
3. When it suggests. Both levels have at least 10 sessions; the cheaper level’s sessions were accepted at least as often, give or take 5 points, even on a cautious estimate; and a typical session at the cheaper level cost at most 85% of one at the current level.
4. What it says. The suggestion names the numbers it rests on, such as At Medium, 12 of its 12 sessions finished with nobody stepping in, against 19 of 20 at High; a typical one cost $0.21 instead of $0.48, and about what it would save a month at the agent’s recent pace.

When either level has fewer than 10 sessions, Spend says Not enough history yet, with how many it has, instead of suggesting anything. When the cheaper level was measured and did worse, or would save little, nothing is shown.

To act on one, as an owner:

1. Open Spend and find Spend less, keep quality.
2. Read the reason under the suggestion.
3. Choose Apply to change the agent’s effort (a new version of the agent, recorded in the audit log with who applied it), or Dismiss to put it away. A dismissed suggestion is not made again for the same setting.

Applied and dismissed suggestions stay listed under Decided lately for 30 days. A suggestion made for a setting that has since changed is put away at the next check.

Budgets

Budgets nest, from the widest to the narrowest. Before every agent reply and session step, g1t checks each level, and the first one that is used up stops the work. Each row on Spend shows the budget, what is spent against it this month, and what happens at 100%.

Level	Set by	At 100%
The workspace	Owners and billing managers, under Billing → Your spend limit.	New work stops until the month turns or an owner raises it: agents, workflows, builds and deploys. With Pause usage at 100% off, owners are alerted and work goes on.
All agents together	Owners, here or under Agents → Budget.	No agent takes new work until the 1st, or until an owner raises it.
Each person	Owners, here. A default for everyone, and a budget of their own for anyone.	Agents take no new work for that person until the 1st, and say so where they were asked.
Each agent	Owners, on the agent’s Profile. A new agent starts with the default set here.	That agent takes no new work until the 1st, or the next day for a daily cap.
Each task	Owners, here: the cap a session starts with. The plan’s caps on one run and one issue are under Billing → Caps.	The session stops at Needs approval, and an owner decides whether it goes on.

A team can also have a budget for what its agents spend together, set on the team’s own settings and checked after a person’s budget. See lead, channel and budget.

Work already running finishes, so spend can go slightly past a budget.

A person’s budget

A person’s budget counts what agents spend on work that person asks for: the replies agents write them and the sessions they start, with everything those sessions bring in. Each person can see their own on You, and in the top bar.

To set budgets, as an owner:

1. Open Spend and find Budgets.
2. Choose Change on All agents together, Each person or Each task, fill in the amounts in dollars, and choose Save budgets. An empty budget means none.
3. To give one person their own budget, choose Set on their row, or Give someone their own budget, and enter it. 0 means no budget at all for them; empty puts them back on the default.

Receipts

Costliest tasks lists the period’s most expensive sessions. Open one for its receipt, at g1t.sh/<workspace>/-/spend/receipts/<session>:

• Each session of its tree: the agent’s own, and every helper and subagent it brought in, with their tokens, the model at the provider’s price, and what each was charged.
• Each session’s model and the effort it ran at.
• Models, at the provider’s price: every session’s model answers together.
• g1t’s agent rate: the rest, at the rate in the price book.
• Total, as budgets count it.

On your own model key, your provider bills the model and only the agent rate is charged; the receipt says so. A session from a conversation you aren’t in shows what it cost, not what it was about.

How it’s priced

The page’s last section reads the price book:

	Price
Models	What the model provider charges, with no markup.
Agent rate	Per million tokens an agent’s run uses, for what g1t runs around every model call: the model gateway, secrets, routing, context and pass-through to your own provider.
Everything else g1t runs	Sandboxes, builds, hosting, storage, search and security scans, at cost plus 20%.
Your own model keys	Your provider bills the model; the agent rate still applies. Local models the same.
Your own runners	$0.
People	No seats.
Spending less	Effort on each agent, and the weekly suggestions above.

In the top bar

Your spend this month sits in the page’s header, beside Ask g1t: what agents did for you, at price, against your budget. Choose it for where it went, by kind of work and by agent, and a link to Spend. Owners and billing managers can switch it to Workspace: everything the workspace used this month at price, the same figure as Home and Spend, with what it was charged after its plan and credit underneath, against its spend limit, and its agents’ share. Both tabs count at price, so they add up: a comped workspace can show $10 used and $0 charged. The choice is remembered on this browser. On a phone, open Spend from the Workspace sidebar instead.

Next

• Effort: how hard each agent works, and what each level costs.
• Agent budgets and spend: who pays for a session tree.
• Usage and billing: the spend limit, caps, AI credit and rates, and how spend is counted.
• Sessions: caps and approving more spend.