Unverified50% confidenceBenchmarkExact time
Under a 4M TPM premise, long Agent turns with a 400K context can trigger at most 10 requests per minute (hitting the TPM cap), and full-context requests only 4 per minute.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
Unverified一个需要20次LLM调用、每次生成200个Token的Agent任务，在30 TPS下需要约133秒完成，在300 TPS下仅需13秒80% similarUnverifiedA short conversation turn consumes about 1300 Tokens/request, in which case the RPM limit (3000) is hit before the TPM limit (4 million); in long-context scenarios TPM becomes the bottleneck.75% similarVerifiedA provider's rate limits typically take effect simultaneously across three dimensions: requests per minute (RPM), tokens per minute (TPM), and tokens per day (TPD).74% similarUnverifiedIn concurrent scenarios, generation speed is evenly distributed among multiple requests; for example, at 40 token/s for a single request, each of 4 concurrent requests gets approximately 10 token/s74% similarUnverified一个Agent的总Token消耗可能达到50K-100K，而同样的对话在Chat模式下可能只需要5K72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1034047API
curl https://kongchang.com/api/v1/knowledge/claims/1034047MCP
get_claim(id=1034047)