Unverified50% confidenceFactExact time
Antirion's billing model reserves a token estimate at admission and settles exactly once with actual usage at request end; when a client disconnects midway the request is not cancelled and the partial cost already generated is still charged.
1
Sources
50%
Confidence
Long-term
Relevance
10/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPer-request billing is flat and predictable; per-token billing is a function of input, output, and the number of loop retries, so the cost shape is different.76% similarUnverifiedWhen a streaming response is interrupted midway, the portion of output tokens already generated is still billed, even though the client may not have received the complete result.75% similarUnverifiedIn streaming responses, the client may disconnect midway or stop reading, so cost recording should be placed in the finally block of try/finally rather than after the consumption loop, to ensure that early exits are also recorded.73% similarVerifiedThinking tokens for reasoning models are typically billed, and setting a token budget cap can prevent runaway costs from a single request.73% similarUnverified给顾问模型设置最大输出上限（如 1000 Token）能同时有效控制成本和等待时延73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1062956API
curl https://kongchang.com/api/v1/knowledge/claims/1062956MCP
get_claim(id=1062956)