Live model pulse

Which model answers stand out in real runs?

A running, anonymized signal from consens.io: every time the independent judge identifies the strongest individual answer behind a consensus, that provider family earns one selection.

Useful signal, honest limits.

This is not a popularity vote and it is not an accuracy benchmark. It shows which answers the judge finds most useful across the changing mix of questions people actually ask.

All-time ranking

Best-answer selections

Provider and product aliases are combined into one family, so Claude and Anthropic never appear as separate competitors.

Kimi and GLM joined consens.io on 31 August 2026, Meta Muse on 2 September 2026. Use the shared-window view to compare the families over one period; every late arrival carries its own start date.

01 OpenAI / ChatGPT 365picks
02 xAI / Grok 225picks
03 Anthropic / Claude 181picks
04 Google / Gemini 179picks
05 DeepSeek 127picks
06 Mistral 76picks
07 Meta / Musetracked since 17picks
08 Moonshot AI / Kimitracked since 15picks
09 Z.ai / GLMtracked since 13picks

1,198 judge selections from all real runs. No prompt text or user identity is shown in this tally.

Two different lenses

Live behavior and controlled accuracy answer different questions.

Model pulse

What stands out in everyday runs?

Questions
Real, changing prompts
Signal
Independent judge selections
Best used for
Product behavior and model momentum
MMLU-Pro benchmark

What is correct under fixed conditions?

Questions
314 known-answer exam items
Signal
Measured accuracy
Best used for
Controlled model and consensus comparison
Open the full benchmark ›

Run your own comparison

Ask once. See every answer. Keep the consensus.

Open app