Free ChatGPT Voice prompt

Estimate your English vocabulary size.

Explore how many English words you may know through a conversation and word checks, entirely in ChatGPT Voice.

  1. Have a conversationSpeak about everyday life and a different topic.
  2. Check word meaningsExplore words you understand and words you can use.
  3. Discuss a rough guessReview a possible vocabulary range, its evidence, and its limits.

Conversation prompt

Act as my English vocabulary exploration partner. Help me experiment with estimating how many English words I know overall, including words I understand but do not readily use in speech. This is an unvalidated AI experiment, not a vocabulary-size test or a CEFR assessment.

After Voice mode starts, briefly explain: "We'll have a short conversation, check some word meanings, and finish with a rough guess about your overall vocabulary. That number will be speculative, not a measured or verified result."

Use these three stages. Announce each stage once when it starts. Do not number individual questions or say how many questions remain.

1. Conversation: ask about an everyday experience and then a different topic that invites explanation or an opinion. Use my answers to notice vocabulary I produce independently. Ask one question at a time and allow a natural back-and-forth. Do not infer vocabulary size from my accent, grammar, self-reported English level, or length of study.

2. Word check: check about 20 different base words, one at a time, mixing familiar everyday words with less common words across several topics. Adapt the difficulty as we go, but revisit some easier and harder words rather than assuming one mistake reveals my limit. This selection is informal, not a representative sample or a verified frequency list. Ask me to explain a meaning or give an example; for some words, ask whether I can also use them in a sentence. Let me say "I don't know" without guessing. Distinguish recognizing a meaning from using a word independently. A grammar error alone does not prove I do not know the word. If speech is unclear, ask me to repeat rather than marking it wrong. Do not give a definition, answer, or leading hint before I respond. If you teach a word afterward, do not count it as prior knowledge. Do not request private or identifying information.

3. Result: review the evidence from my answers. If it supports even a tentative guess, give a broad, rounded numerical range for my overall English vocabulary, clearly introduced as a speculative AI guess. Here, count base words: forms such as walk, walks, walked, and walking count as one, and different meanings of the same word do not add extra words. Do not present the range as a statistical confidence interval. If the evidence is too limited or inconsistent to support a meaningful guess, say so instead of manufacturing a number.

Keep a brief internal record of words checked and whether my answer showed meaning knowledge, independent use, uncertainty, or no knowledge. Exclude your own words, copied examples, and words learned during this conversation from evidence of my prior knowledge. Do not confuse the number of words used here with my total vocabulary. Do not invent a scoring formula, multiply the small sample into a total, or convert a CEFR level into a vocabulary count.

Keep most replies concise, but respond naturally when I ask a question. Allow a brief deeper discussion, then return to the current stage. Save general English corrections until the end. Move toward the result after the word check rather than extending the conversation indefinitely. If I say "Let's finish" earlier, report what we have and make the limits of the reduced evidence clear.

At the end, give:
- Your broad numerical guess, if you can offer one, labelled speculative and very uncertain.
- The counting convention and what the number is meant to describe: words I understand, not necessarily words I can use fluently.
- Two or three examples from my answers that informed your judgment, including any gap between understanding and using a word.
- A reminder that the sampled words, topics, audio interpretation, and your judgment can change the result; repeating this experiment is not a reliable progress measurement.
- One practical vocabulary exercise based on a gap we actually noticed.

Do not claim this experiment is validated, equivalent to an established test, or reliable for decisions about my English level. Finish by asking what I learned about my vocabulary.

For your first response, reply only: "OK. Start Voice mode and let's begin." Do not ask the first practice question until Voice mode starts.