Free · no account required
Free TCF Canada listening practice
Practise with 39 original French recordings and explained answers. Free short sets, a 35-minute listening exercise, transcripts and missed-question retries.
39 original French audio questions. All practice, transcripts, explanations and missed-question retries are free without an account. These are independent fictional exercises with synthetic narration; they are not official or recalled questions.
Format, audio and saved progress
FEI lists 39 listening questions in 35 minutes, four choices and one correct answer. Each official recording is played once and the question follows the recording. The complete TCF Canada also includes reading, writing and speaking.
Our timed exercise uses 39 questions and 35 minutes, with one playback per completed item. You control when to move to the next item; this is not the official continuous test player. Our audio-only items do not reproduce official picture questions or calibrated difficulty. An interrupted recording can be restarted and the attempt is then labelled as assisted.
Progress is saved only in this browser, including the original timed deadline. Reloading or changing interface language does not restart the timer. Audio files are fetched from this site when needed. No microphone, speech upload or model call is used; one current attempt is retained. Storage failure is shown and does not block practice.
Official FEI listening format (French)Preparing saved progress…
Choose your listening practice
Replay as needed, choose an answer, then check the explanation and transcript. Use the short sets first. In study mode the question is visible before listening; there is no timer.
13 questions · The three 13-question sets and the timed exercise reuse the same 39 recordings; they are not additional banks.
How to use the original listening bank
Listen for the speaker’s actual purpose, time changes and conditions. Choose the option supported by what is said, rather than one that repeats a familiar word. After checking, compare your choice with the short evidence quote and listen again in study mode.
Short messages · 13
Everyday announcements, changed plans and instructions. Separate the original arrangement from the new request.
Details and conditions · 13
Connect several sentences and preserve limits: a proposal is not an automatic confirmation, and an estimate is not a guarantee.
Viewpoints and purpose · 13
Identify the speaker’s position, qualification or purpose. A cautious proposal should not become an absolute claim. These are teaching groups, not official sections or CEFR bands.
The three 13-question sets and the timed exercise reuse the same 39 recordings; they are not additional banks. The count and percentage describe these original items only. They do not convert to an official TCF score, CEFR or NCLC level and do not predict your exam result.
TCF Canada listening questions
Are the 39 questions official exam questions?
No. The French scripts, questions, choices and explanations are original fictional learning material. Narration is generated offline using a disclosed synthetic French voice. Official examples remain on FEI’s website; we do not use recalled questions.
How does the timed exercise work?
It uses this bank of 39 questions and a 35-minute deadline. Each completed recording plays once; you choose when to move on. Written choices unlock after the spoken question. Reloading keeps the original deadline. Restarting interrupted audio is permitted and labelled as assisted. This is an independent listening exercise, not the complete official four-skill exam.
Can I replay, read transcripts and retry mistakes for free?
Yes. Study mode permits replay and shows the explained answer and French transcript after you check. Timed mode withholds corrections until the review. The results page lets you retry incorrect and unanswered items as a new untimed set. No account or paid unlock is required.
Can my percentage give me a TCF or NCLC score?
No. The bank is not calibrated to official difficulty or score scales. Your correct count and percentage describe only this exercise. Repeated attempts reuse the same recordings and may benefit from memory.
Audio source and licence
Original scripts are narrated with Piper / Tom (French, France), generated offline using the unmodified AGPL-3.0 model source. These are single-voice narrated passages, not human recordings or a range of regional accents. No voice model or TTS runtime is shipped to visitors. The voice source, upstream project and per-clip provenance are retained.
Piper · Tom · AGPL-3.0