Expired70% confidenceFactExact time
Sonnet 5.5's cache read cost remained at $0.20 per million tokens with no reduction, whereas Fable 5.1 cut it by 90% and Opus cut it by approximately 60%.
4
Sources
70%
Confidence
Short-term (~14 days)
Relevance
10/2/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/16/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedIn agentic coding work, the proportion of cache reads in total cost is under 5% in Fable, close to 20% in Opus, and potentially over 50% in Sonnet 5.5.77% similarUnverifiedOn the codebase audit benchmark task, Sonnet's cost was approximately half that of Opus but scored slightly higher, achieving the best cost-per-point, and Sonnet took only about 5 minutes while Opus took nearly twice as long.68% similarUnverifiedAnthropic's input/output token pricing for Fable 5.1 remains the same as Fable 5, but the cost of cached token usage has been significantly reduced.68% similarUnverifiedThe core selling point of Sonnet 5.5 is faster response speed and lower token consumption.68% similarUnverifiedSonnet 5 is not a token-efficient model; it consumes a large number of extra tokens on agentic tasks, causing the actual API cost to be higher than that of Opus 4.866% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/967377API
curl https://kongchang.com/api/v1/knowledge/claims/967377MCP
get_claim(id=967377)