BNKR BankrCoin
BNKR on Base is the Bankr platform token. Product documentation establishes its AI-agent role; launch documentation describes fees allocated to buybacks.
Editorial focus · AI / DeFi
AI agents
A token funds or represents an agent. Look for a customer, a paid task and a useful result.
WatchingMechanism described by the teamSources checked:
Build the current meta
Object
An agent or service
Promise
Useful automation
How buyers are reached
Agent directory and social launches
Metric
Paid tasks less compute costs
Liquidity source
Token buyers and people supplying trading liquidity; customer payments are a separate source
What if the buyer is an AI agent?
Data request → service asks for payment → agent checks its limit → USDC payment on Base → service response. This is an x402 flow; the protocol also works on other networks.
Where does the human stay in control? +
Before signing, the client checks the network, asset, recipient and spending limit. A request over budget should not be paid. This explains the protocol; it is not a Yieldo payment service.
Market examples
Check the asset’s role and who receives the money: distribution rules differ between projects.
↳
AI agents
Who gets paid
The token buyer, product customer and fee recipient may be different people. The diagram shows what each pays for and receives under the project’s rules.
- 01Payer
AI service customer
- 02Action
Orders a completed task
- 03Payment or fee
Task payment; trading fees separately
- 04Recipient
Service, operators and compute providers
- 05What the holder receives
Access or distributions only under the specific token's rules
Money in the market
Token buyers and trading liquidity providers.
Money from use
Customer budgets for completed tasks net of direct execution costs.
Who can earn while the token falls
Services and compute providers can get paid without token appreciation.
What would disprove the promise+
Every strategy relies on a condition. This section explains what must work and what result would make the bet worth reconsidering.
- Testable promise
- The agent completes tasks customers will repeatedly pay for.
- Evidence against the hypothesis
- Token trading grows but repeat paid tasks are absent, or direct task costs persistently exceed payments.
- How to check
- Check task output, repeat orders and payments minus compute. Track free calls and token volume separately.
- Observation window
- Two 30-day windows with a consistent service sample and definition of a paid task.
Mechanism profile
- What was tokenized
- AI infrastructure, an agent, service access or a fundraising asset, depending on the project.
- The promise
- Automation and early access to autonomous income.
- The proposed new buyer
- A customer buying automation, or an agent with a budget funded from outside the token economy.
- How to check demand
- Repeat paid tasks after computing costs. Agent counts and token market value are not customer counts.
- What worked
- Automation and interfaces for specific tasks where customers valued the result.
- Where value becomes misleading
- An AI content account was treated as a business; token trading fees were confused with AI service sales, and circular payments with outside demand.
- Inherited mechanism
- Revenue promises, computing financed by tokens and automated asset creation.
What to check before participating
- Reproduce a useful task and identify the paying customer.
- Separate service sales from the token’s trading fees.
- Compare payments with computing costs and look for repeat orders.
Sources and verification scope
Source review dated 24 September 2026. A map of mechanisms, not a yield ranking.