count tokens

This block is part of the anthropic group and was last modified in core v15.0.0.

count tokens anthropicconnection v15.0.0

Count the input tokens a prompt (with system + attachments) would use, and estimate the cost; before you send it. Free; no spend.

In pins 6 + flow

flow
flow
connection required
anthropicconnection
prompt
value
The user message to measure.
system
value
Optional system prompt to include in the count.
attachments
file
Optional images/PDFs; their tokens are counted by the API too.
model
value
Model to price against. Defaults to the connection's default model. • Suggestions: claude-opus-4-8 claude-sonnet-4-6 claude-haiku-4-5
max tokens
number
Max output tokens, for the worst-case output cost estimate. Default 16000.

Out pins 11 + flow

flow
flow
connection
anthropicconnection
success
condition
model
value
input tokens
number
Tokens the prompt + system + attachments would use.
max tokens
number
The max output tokens used for the output-cost estimate.
input cost (USD)
number
Cost of the counted input tokens.
max output cost (USD)
number
Worst-case output cost if generation used the full max_tokens.
estimated total (USD)
number
input cost + max output cost (the upper bound for a single call).
errors
array<value>
Error codes when the call fails (success=false): config_error, rate_limited, transport_error.
warnings
array<value>
Non-fatal notices: cost_unknown (model not in the price table; costs reported as 0).
error detail
value

token count

estimate cost

tokens

price check

how much

A typical wiring for count tokens: a function trigger starts the flow; the flow out pin feeds debug.

Studio canvas example for the count tokens block: typical wiring for count tokens.

Introduced in v15.0.0.

Cluster