Parakeet TDT 0.6B v3 vs Voxtral Mini 3B
Two open-weight speech-to-text models on the same endpoint and API key, so switching between them is a one-word change. Below: what each costs for three realistic workloads, how they differ, and measured speed.
Inference APIs prices are read live from the price list.
- Cost: Voxtral Mini 3B is cheaper in all three workloads, by 35%.
Side by side
| Parakeet TDT 0.6B v3 | Voxtral Mini 3B | |
|---|---|---|
| Served by | Inference APIs | Inference APIs |
| Model author | NVIDIA | Mistral AI |
| Model id | nvidia/parakeet-tdt-0.6b-v3 | mistralai/Voxtral-Mini-3B-2507 |
| Price | $0.002 / audio minute | $0.0013 / audio minute |
| Capabilities | Speech to text, Very fast, Timestamps, 25 languages | Speech to text, 8 languages, Auto language, Lowest price |
| Weights | Open | Open |
| Status | Available | Available |
| Limits | No per-minute or per-day request or token caps; usage draws on a prepaid balance | No per-minute or per-day request or token caps; usage draws on a prepaid balance |
| Speed | 0.036× real time (19.5 s clip in 0.7 s) | 0.058× real time (23.1 s clip in 1.35 s) |
What three workloads cost
List prices applied to the same work. The cheaper side of each row is in bold.
| Workload | Parakeet TDT 0.6B v3 | Voxtral Mini 3B | Difference |
|---|---|---|---|
| 1 hour of audio | $0.12 | $0.08 | 35% |
| 100 hours of audio | $12.00 | $7.80 | 35% |
| 1,000 hours of audio | $120 | $78.00 | 35% |
Voxtral Mini 3B is cheaper in all three workloads, by 35%.
When to choose which
- Long recordings in one of its 25 European languages, where throughput matters
- Difficult audio: Voxtral Small or Whisper are more accurate
- Timestamps, subtitles or diarization
- Cost is the deciding factor; it is cheaper in every workload above
- Cheap, fast multilingual transcription in Voxtral's eight languages
- Batch jobs where cost per minute matters most
- Your audio is in a language outside its 25; use Whisper
- You need live streaming transcription or speaker diarization
Measured speed
Median of three runs on 2026-09-16: a 19.5-second mp3 clip, timed from upload to response. Expect ±30% with time of day. Method and raw numbers: speed measurements.
Trying both
Same endpoint, same key. Change one string:
for MODEL in "nvidia/parakeet-tdt-0.6b-v3" "mistralai/Voxtral-Mini-3B-2507"; do
curl -s https://api.inferenceapis.com/v1/audio/transcriptions \
-H "Authorization: Bearer $INFERENCE_API_KEY" \
-F "model=$MODEL" -F "file=@meeting.mp3" | jq -r .text
doneFAQ
Which is cheaper, Parakeet TDT 0.6B v3 or Voxtral Mini 3B?
Voxtral Mini 3B is cheaper in all three workloads, by 35%. The table above uses list prices: Parakeet TDT 0.6B v3 $0.002 / audio minute; Voxtral Mini 3B $0.0013 / audio minute.
How do I switch between them?
They are on the same endpoint and key. Change the model field from nvidia/parakeet-tdt-0.6b-v3 to mistralai/Voxtral-Mini-3B-2507 and nothing else.
Spotted a price or limit that has changed? Tell us and we will re-check the provider page.