Unverified50% confidenceBenchmarkExact time
在LELP-S+测试中，GPT的Token节省率为44%，Gemini为36%，Claude为32%，DeepSeek为30%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/11/2026
First Seen
Valid until: 11/9/2026
Sources
Related Claims
Partially VerifiedClaude Opus 4.8 scored 69.2% on the SWE-bench benchmark, compared to 58.6% for GPT 5.5 and 54.2% for Google Gemini.68% similarUnverifiedThe median output token count per problem-solving attempt for Opus 4.7 is 60,000, while GPT 5.5 requires only 16,000.67% similarUnverifiedOn mainstream benchmarks such as MMLU and HumanEval, the score gaps among Claude 3.5, GPT-4o, Gemini 1.5 Pro, and Grok-2 have been compressed to within the range of statistical error.66% similarUnverifiedOn the DeepSWE benchmark, GPT-5.6 scored 45 at Low, 61 at Medium, 69 at High, 71 at X-High, and 73 at Max across reasoning tiers.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/730306API
curl https://kongchang.com/api/v1/knowledge/claims/730306MCP
get_claim(id=730306)