Unverified50% confidenceFactExact time
Prompt token reading can be processed in batches in parallel, but output generation can only proceed token by token, so the read rate and the write rate are two different numbers.
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedThe computational cost of generating output is typically far higher than that of processing input, and some vendors set separate quotas for input tokens and output tokens.70% similarUnverifiedA piece of code is read far more times than it is written.67% similarUnverifiedThe cached input volume is far higher than the output volume because each self-check round requires feeding a large amount of existing context back into the model.66% similarUnverified投机解码主要优化TPOT（每Token生成耗时）指标，在低并发Memory-bound场景下将闲置算力转化为草稿验证的并行计算，单用户速率提升幅度显著66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/977207API
curl https://kongchang.com/api/v1/knowledge/claims/977207MCP
get_claim(id=977207)