Unverified50% confidenceFactExact time
Input tokens can be processed in parallel with high utilization; output tokens must be generated one by one autoregressively, occupying more memory bandwidth and being slower, so most providers set the output-side unit price significantly higher than the input side, with a common ratio of 3:1 to 5:1.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOutput token pricing is typically 2–4 times that of input tokens, because the generation process requires step-by-step decoding and computation of attention weights.76% similarUnverifiedOutput tokens are typically more expensive than input tokens because outputs require autoregressive token-by-token generation, which has higher computational intensity and lower GPU utilization.74% similarVerifiedThe price of output tokens for large language models is typically 4–8 times that of input tokens, because output involves autoregressive generation requiring multiple forward passes, resulting in higher marginal costs.73% similarUnverifiedThe 4.8x figure refers to the ratio of total tokens processed per GPU per second at the point where single-user response speed is identical, not that answers are 4.8x faster or that 4.8x more users can be served simultaneously73% similarVerifiedThe price of output tokens is typically 4–6 times that of input tokens, stemming from the computational differences between the prefill and decode phases, with GPU utilization during the decode phase potentially being only 10–20% of that during prefill72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1005632API
curl https://kongchang.com/api/v1/knowledge/claims/1005632MCP
get_claim(id=1005632)