models.fyi

"It got dumber" — now with evidence

The model behind an endpoint changes: silent checkpoint swaps, quantization changes under load, serving-stack regressions. We re-run the same probe suite on the same configurations every week, so "降智" stops being a feeling and becomes a number with a date. Latest run: Jul 6, 2026.

GPT-5.5 · OpenAI · MediumGLM-5.2 · DeepInfra · MediumGemini 3 Pro · Google AI · HighKimi K2 Thinking · Groq · HighClaude Fable 5 · Anthropic · High

Weekly Intelligence Index · dashed rules mark serving-change events

Attributed serving changes

Jun 7, 2026Kimi K2 Thinking · Groq-1.5 ptsstill in effect

Speculative-decoding aggressiveness increased for throughput; quality dip still in effect.

May 31, 2026GLM-5.2 · DeepInfra-2.4 ptsrecovered Jun 21, 2026

Quantization silently switched int8 → int4 during a capacity crunch; reverted after three weeks of community reports. The canonical "降智" case.

Apr 27, 2026GPT-5.5 · OpenAIMedium tier only-1.1 ptsrecovered May 25, 2026

Serving-stack regression in medium-effort routing under load; acknowledged by OpenAI and fixed four weeks later.

Mar 8, 2026Gemini 3 Pro · Google AI+0.9 ptsstill in effect

Silent checkpoint update — math and coding scores improved without an announcement. Drift cuts both ways.

Feb 15, 2026DeepSeek-V3.2 · DeepSeek (official)-0.7 ptsrecovered Mar 1, 2026

Chat-endpoint prompt-template bug degraded instruction following for two weeks.

Recent shifts

ConfigurationΔ index
Kimi K2 Thinking · Groq-1.2 pts
GLM-5.2 · DeepInfra+0.6 pts
GPT-5.5 · OpenAI+0.2 pts
GLM-5.2 · Fireworks+0.1 pts
GLM-5.2 · SiliconFlow+0.1 pts
Claude Opus 4.8 · Anthropic-0.1 pts
Claude Fable 5 · Anthropic+0.1 pts
GLM-5.2 · Z.ai (official)+0.1 pts

Method: mean of the last 4 weekly runs minus the mean of the 4 before, per configuration at its default effort. Run-to-run noise is ±0.15 index points; we only attribute an event above ±0.8.