Three steps from recording to real feedback.
I do not run an app or a multi-week course. Send a sample today, I send back a one-paragraph diagnosis under the 2026 TOEFL rubric within 3 days, then you send the next one. That is the whole loop.
Submit a sample
Record a 60-second Speaking response, or paste a Writing draft. Use prompts from ETS, TPO, my Speaking course, Writing course, YouTube, or any TOEFL textbook.
- One sample per 24 hours
- Speaking: audio link (Drive, Dropbox, OneDrive)
- Writing: paste the draft into the form
Receive personal feedback
I read every sample personally and grade it the way an ETS rater would. For Speaking samples, I send back an audio recording of me grading your sample live, you hear my reaction, suggestions, and score in real time. For Writing samples, a one-paragraph diagnosis. Either way you get an estimated score on the 0 to 30 (Old TOEFL) or 0 to 6 (2026 TOEFL) scale, plus the two or three things costing you points most.
- Audio grading recording (Speaking) or written diagnosis (Writing)
- Estimated score under the current rubric
- Specific weaknesses, in plain language
- Specific fixes you can apply on the next sample
Apply the fix, then send the next one
The cycle is short and ruthless. You apply the fix from yesterday's diagnosis, record again, and resubmit. Most students plateau because they keep practicing the same mistakes. The diagnosis tells you exactly what to break out of.
- Track progress across samples
- Build muscle memory for the rubric
- Walk into test day knowing what raters reward
What this is not.
This is not an automated AI grader. This is not a course you buy upfront. This is not a guarantee of any particular score. It is one university English professor reading your work and telling you, in plain language, what is and is not working under the 2026 rubric.
If you want the full subscription with daily samples and Zoom office hours, see pricing. The first sample is always free.
Ready to send your first sample?
Send a free sample ›