Narration and explainers
Long reads where an even pace matters, the register our note describes. Word-level captions come with every job.
Elodie, the one French voice on the list, with a measured delivery by our note
Elodie is the only French voice on the published list, and the best-graded voice outside English: a B- with B target quality and under eleven hours of training audio from a public French speech corpus. Our note is a listening note, measured delivery, and two recipes we kept start from her, with Delaney and with Maya.
FrenchFemaleB-Word-level captionsStandard voice
Hear it
Three samples, three registers
- Listen for the measured pace our note describes over the long harbour sentence.
- Listen for whether an unhurried voice can carry the urgency of a ten-second hook.
- Listen for the pause before honnêtement, and how the aside sits against the sentence before it.
On the record
A B- from under eleven hours, and a note about pace
Under eleven hours puts her in the same band as Sakura and above every Hindi and Italian voice on the list, and her B- is the best grade outside English. Target quality B means the corpus recordings were reasonably clean. The grade measures the data, not the sound.
Measured delivery is our note, and it is about pace: an even, unhurried read. The two recipes we kept keep her at 30 percent, so her French pronunciation stays in charge while Delaney or Maya at 70 percent changes the colour.
Length the engine likes. The published guidance puts the sweet spot at roughly 100 to 200 tokens, a few sentences; very short lines are the weak spot and very long ones can rush.
Where we would use it
Narration, explainers, and French with a borrowed timbre
French product and course videos
Scripts read at one steady pace; tag any name that must sound a particular way with Pronounce as.
The base of two kept recipes
Husky French and Clear French both keep her at 30 percent. They are in the generator presets.
Blends
Two kept recipes that keep the French at thirty percent
The Delaney and Maya cards are recipes we listened to and kept; the others are pairings to try. Elodie keeps the French pronunciation in each.
DelaneyEnglish (US)
With Delaney
Husky French, kept at 70 percent Delaney: a French read with a throaty edge, by our note.
70% Delaney
MayaEnglish (US)
With Maya
Clear French, kept at 70 percent Maya: a French read with Maya's clarity, by our note.
70% Maya
ImogenEnglish (UK)
With Imogen
A pairing to try. Imogen is the top-graded British voice with a courseware note; listen for a courseware timbre on French.
60% Imogen
GiuliaItalian
With Giulia
A pairing to try, two Romance-language voices: Giulia is our Italian short-form voice by our note.
50% Giulia
Recipes we kept with Elodie
Husky French
French read with a throaty edge.
Elodie 30% + Delaney 70%
Clear French
French read with Maya's clarity.
Elodie 30% + Maya 70%
Blend Elodie with any second voice
A fixed sample at this mix, rendered once and cached. No credits used.
A cross-language blend: Elodie keeps the pronunciation, the second voice only lends its timbre.
Every blend works the same way: the first voice sets the pronunciation and pacing, the second voice lends only its timbre, and you choose the ratio between 10 and 90 percent. See the three steps on the voice library page
At a glance
Elodie on the published list
| Language | French |
|---|---|
| Voice | Female |
| Overall grade | B- |
| Target quality | B |
| Training audio | 1 to 10 hours |
| Captions | Word-level captions |
| Cost | 1 credit per job, room for 10,000 characters |
| Formats | MP3 free, WAV on paid plans |
| Commercial use | Included on every plan |
| Preview | Free, no sign-up |
Grade, target quality, and training audio come from the engine's published voice list and estimate training data, not how a voice sounds.
FAQ
Questions about Elodie
Why is Elodie graded higher than the Hindi and Italian voices?
Training audio. Under eleven hours is far more than their tens of minutes, and the published grade tracks the amount and quality of data, not how a voice sounds.
Do the captions follow each word in French?
Yes. French returns word-level timing, so the SRT and WebVTT can highlight word by word as well as line by line.
Why do the kept recipes keep her at only 30 percent?
Because the first voice controls pronunciation and pacing regardless of share. At 30 percent she still reads every word; the second voice at 70 percent only lends its timbre.
Is there a French landing page?
Not yet. Use the main generator, which accepts French text directly, or start from this page with the Use Elodie button.
Generate with Elodie
Paste a script, keep Elodie or blend a second voice, and download the MP3 with subtitles.