Unverified50% confidenceBenchmarkExact time
Opus 5.5 is less token-efficient than Opus 5: Opus 5's max tier uses 73k tokens per task, Fable uses 78k, Opus 5.5's max tier approaches 120k tokens, while GPT-6 Astra only requires 27k.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/25/2026
First Seen
Valid until: 12/24/2026
Sources
Related Entities
Related Claims
UnverifiedThe median output token count per problem-solving attempt for Opus 4.7 is 60,000, while GPT 5.5 requires only 16,000.78% similarUnverifiedOpus 5's pass rate at Low Effort is approximately 60%, surpassing Opus 4.8's 59% at the Max level, for which the per-task cost is as high as $13.73% similarUnverifiedOn deep code review tasks, Grok 4.7 and Sol outperform Opus and Fable, but Opus's improvement on that bench from 5 to 5.5 is close to doubling.70% similarUnverifiedOpus 5.5 is the first version with targeted improvements to sentence clarity and information overload, but the engineer emphasized this does not mean Opus 5.5 has comprehensively surpassed Opus 4.6.69% similarUnverifiedAccording to benchmarks from Artificial Analysis and Cursor, GPT-5.6 Sol requires approximately 17k–28k tokens to complete a task, while Anthropic's Fable requires 36k to 103k tokens.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/948924API
curl https://kongchang.com/api/v1/knowledge/claims/948924MCP
get_claim(id=948924)