Flashcard audio: the word and its translation, one MP3 per card
Open a CSV with words in one column and translations in the other, or paste the two columns from a spreadsheet. Speecho reads each word in a native voice of its language, pauses, then reads the translation in a voice of the other language. You get one MP3 per card, named after its word, in one ZIP, and a CSV that Anki imports with the sound on every card.
Voice your flashcards →From a word list to cards that speak
Two voices, one for each language, and the same pause on every card.
- 0001_lopen.mp3
- 0002_maken.mp3
- 0003_spelen.mp3
- 0004_Hoe_gaat_het.mp3
- …
- 0200_kinderen.mp3
Who voices cards
Anywhere a word should be heard together with what it means.
Anki decks
Cards with text only, or audio recorded and attached word by word.
Now: one sound per card, and a CSV that imports the deck with the sound in place.
Language teachers
A vocabulary list for every lesson, read aloud in class or recorded at home.
Now: the list as audio cards to share, the word in the target language, the meaning in the students' own.
Language apps and games
Hundreds of word and translation pairs, each needing its own audio asset.
Now: one ZIP with a file per pair, numbered in the list's order and named after its word.
Travel phrasebooks
Phrases nobody knows how to pronounce, next to a written translation.
Now: each phrase said by a native voice, then its meaning, in one short file.
Exam vocabulary
Long word lists to learn on the way to school or work.
Now: a folder of cards to play in order on any phone, eyes free.
Children learning a second language
Picture cards with a word nobody at home can say well.
Now: the word spoken by a native voice, then in the child's own language.
A card set at a glance
- 2 voices per card: one for the word, one for the translation
- 1 MP3 per card, or two files if you want the sides apart
- 1,000 cards in one set, and up to 200,000 characters
- 10 characters: the least a side is billed as, so a word is never priced like a paragraph
- 3 pauses to choose from before the translation: 0.5, 1 or 2 seconds
- 1 CSV for Anki beside the ZIP, its sound fields naming the files
What a card set costs
All the words and translations are priced together as one text by their characters, with a single 10-credit minimum for the whole set. The price is on the button before anything is charged.
Packs: $5 = 2,400 credits, $10 = 6,000, $25 = 16,800, $50 = 36,000. Credits never expire, there is no subscription, and there is a 14-day refund. All prices on the pricing page →
Frequently asked questions
How do I add audio to Anki flashcards?
Voice your cards in Speecho and download the ZIP and the Anki CSV. Unzip the files into the collection.media folder of your Anki profile, then import the CSV as notes: the third column holds the sound as [sound:0001_lopen.mp3], so every card plays its word and translation.
Can the word and the translation have different voices?
Yes, that is the point. The words get one voice and the translations another, each picked in its own language: a Dutch voice for "lopen", an English one for "to walk". The language of the translations is read from them, and you can change it.
One file per card or two?
One by default: the word, a pause and the translation in one MP3, which is what most flashcard apps want. Choose two per card to get the word and the translation as separate files, 0001_lopen.mp3 and 0001_lopen_2.mp3, for the front and the back.
Which file format do I need?
A .csv or .tsv with the word in the first column and the translation in the second, as any spreadsheet saves it; a header row is left out. You can also paste two columns from a spreadsheet, or type each card as word → translation.
How much does it cost?
Both sides of every card are priced together as one text: 50, 110 or 160 credits per 1,000 characters on Standard, Enhanced or Premium voices, with one 10-credit minimum for the whole set and every side counted as at least 10 characters. A thousand short cards come to about 1,000 credits on Standard and 3,200 on Premium.
How many cards can one set have?
Up to 1,000 cards and 200,000 characters on both sides together. A longer deck goes in as several sets, each with its own ZIP and CSV.
What if a word is pronounced wrong?
Open the set in your history, tick the side and voice it again: every take comes out a little different. You can also correct the text first, for example add the article or a full stop. A side voiced again costs only its own characters.
Can the translation come first?
Put the translation in the first column and the word in the second: the first column is always read first. The voices follow the columns, so pick them the other way round.
Which languages can I use?
The same languages as Text to speech: native voices for many languages and accents, and universal voices for the rest. Each side can be in any of them.
Can I use the files commercially?
Yes. Generation runs on the paid commercial API tiers of our voice providers, whose terms permit commercial use of the generated audio. You are responsible for the content of your cards.
Other ways to turn text into audio
- CSV to Anki deck → The ZIP and the Anki CSV as one .apkg deck, sound already in place.
- Quizlet to Anki converter → A Quizlet export or any word list as a file Anki imports, free, in your browser.
- Bulk text to speech → A word list in one voice, an MP3 per line.
- Text to speech → Any text in one natural voice, as one MP3.
- Multi-voice text to speech → A dialogue with a voice per character, joined into one file.
- Pricing → Every rate and credit pack on one page.
Your vocabulary, as cards that speak.
Open the CSV, pick two voices, download the ZIP and the Anki CSV. A thousand cards on Premium cost about 3,200 credits.
Voice your flashcards →