Certificate verification

Verified recordASK10AIS-2026-000002

全球最受欢迎的电动牙刷品牌

Xiaomi小米

20.12/ 100

Ranked 6th of 21 candidate brands

Certificate No.
ASK10AIS-2026-000002
Issued on
2026-08-24
Evaluated on
2026-08-24
Panel size
12 models

2026-08-2412 modelsask10ais.com/verify/ASK10AIS-2026-000002

This page shows the evaluation snapshot at the time of issuance — it does not reflect the current live ranking. ask10ais keeps running new evaluation rounds; for the latest results, see ask10ais.com.

We do not sell optimisation and we do not promise to raise your recommendation rate. Rankings cannot be changed for money.

The question, verbatim

This is exactly what every model in the panel was asked — no per-model rewording.

全球最受欢迎的电动牙刷品牌

请列出你认为最符合的前 10 个品牌,按受欢迎程度从高到低排序。只输出品牌名称本身,不要加编号或解释。

The fixed suffix every model receives: List the top 10 brands you consider the best fit, ordered from most to least popular. Output brand names only — no numbering, no explanation.

Evidence: per-model scoring

The table below shows the raw rank and contribution each model in the panel gave — you can independently verify how the index score above was computed, not just take a single summary number on faith.

Expand any model to read what it actually returned. The ranks and weights in the table above come from the snapshot frozen at issuance; the raw responses are read from the response log, which is append-only and never rewritten — two records of the same call.

ModelTypeWeightModel's rankContributionRaw responses
×openai/gpt-5.5-proHeadNot counted
tencent/hy3Long-tail16.19%9th0.324
xiaomi/mimo-v2.5Long-tail14.27%Not mentioned0.000
qwen/qwen3.8-maxHead8.61%Not mentioned0.000
z-ai/glm-5.2Long-tail7.42%10th0.074
google/gemini-3.7-flashHead6.94%9th0.139
deepseek/deepseek-v4-proHead6.41%4th0.449
nvidia/nemotron-3-ultra-550b-a55b:freeLong-tail6.31%4th0.442
x-ai/grok-4.6Head5.88%5th0.353
×stealth/ox-alphaLong-tailNot counted
anthropic/claude-opus-5Head3.43%7th0.137
moonshotai/kimi-k3Head1.57%5th0.094

Where this could be wrong

The index level is driven mostly by candidate-pool size, and is not a confidence measure.

MethodologyRefusal log


This record measures what large language models answered to one question at one moment. It is not an assessment of product quality, market share or corporate standing. Model output may contain errors or fabrications; this record preserves the raw responses as returned and does not endorse their content. It is not investment or procurement advice.