Unverified50% confidenceBenchmarkExact time
In the DeepSWE benchmark, the ratio of input tokens to output tokens during agent execution reached 174 to 1.
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn the DeepSWE test, 99.6% of all input tokens were cache hits.72% similarUnverifiedThe ratio of input to output tokens surged from 4.52:1 to 13:1.66% similarUnverifiedThe DeepSuite benchmark has a false positive rate of 44.9% and a false negative rate of 1.2%.58% similarUnverifiedIn the blogger's sample, approximately 94% of Astra's input and approximately 79% of Sonnet's input came from cache.58% similarUnverifiedOfficial tests show that after integrating OpenViking, long-conversation memory accuracy for several agents exceeded 80%, and input token consumption was reduced.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/926821API
curl https://kongchang.com/api/v1/knowledge/claims/926821MCP
get_claim(id=926821)