Unverified50% confidenceFactExact time
Most APIs automatically append conversation template markers before and after the input, and these hidden costs are usually a few dozen tokens.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedMainstream large model APIs charge separately for input tokens and output tokens, with input tokens typically including the entire conversation history (i.e., the context).78% similarVerifiedAPI billing covers both input tokens and output tokens, and as a conversation extends, the input tokens carried in each request grow linearly.72% similarUnverifiedSystem prompts are included in the context and counted toward token consumption in every round of dialogue; system prompts for customized services like Fable may contain hundreds or even thousands of tokens and are not visible to users.71% similarUnverifiedAutomatic prompt tuning consumes a large number of tokens in the model's back-and-forth self-correction process, incurring additional API call, token, and latency overhead.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1017219API
curl https://kongchang.com/api/v1/knowledge/claims/1017219MCP
get_claim(id=1017219)