INTELLIGENCE BRIEFING · ISSUE 003

SENTIENT WEEKLY

Signal in the AI noise.

A weekly editorial briefing · free in your inbox

— Benchmarks —
GPT-5.5 (xhigh)64.8
GPT-5.5 (medium)61.0
Claude Opus 4.7 (Non-reasoning, High Effort)58.1
Gemini 3.1 Pro Preview57.6
GPT-5.5 (low)52.2
MiMo-V2.5-Pro51.2
Muse Spark50.9
DeepSeek V4 Pro (Reasoning, High Effort)50.1
DeepSeek V4 Flash (Reasoning, Max Effort)48.3
GPT-5.4 mini (xhigh)48.0
GPT-5.5 (xhigh)64.8
GPT-5.5 (medium)61.0
Claude Opus 4.7 (Non-reasoning, High Effort)58.1
Gemini 3.1 Pro Preview57.6
GPT-5.5 (low)52.2
MiMo-V2.5-Pro51.2
Muse Spark50.9
DeepSeek V4 Pro (Reasoning, High Effort)50.1
DeepSeek V4 Flash (Reasoning, Max Effort)48.3
GPT-5.4 mini (xhigh)48.0