modelbenchmark.io

Voxtral Small 24B

Mistral · voxtral-small-24b

Compare

Mistral Voxtral audio-understanding model for speech and transcription tasks

Specification

most-agreed values

Context
33K
Max output
26K
Released
2025-07-15
Knowledge cutoff
2025-07
Retires
2026-08-01
Open weights
yes
Input
text, audio
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.10
Output
$0.30
Cache read
$0.01

Available from 8 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
evroc$0.0023$0.002332K32K
amazon-bedrock$0.10$0.3033K8K
bedrock_converse$0.10$0.30128K8K
edenai$0.10$0.30128K33K
kilo$0.10$0.30$0.0133K26K
openrouter$0.10$0.30$0.0133K26K
scaleway · scaleway$0.15$0.3532K16K2026-08-01
greenpt$0.228$0.51333K16K

Price history

append-only observations · a listing writes a row only when its price moves

DateHostField$/M
2026-09-02openrouterCache read$0.01
2026-09-06openrouterCache read$0.01
2026-09-02amazon-bedrockInput$0.15
2026-09-12amazon-bedrockInput$0.10
2026-09-02openrouterInput$0.10
2026-09-06openrouterInput$0.10
2026-09-02amazon-bedrockOutput$0.35
2026-09-12amazon-bedrockOutput$0.30
2026-09-02openrouterOutput$0.30
2026-09-06openrouterOutput$0.30

More from Mistral

most-hosted first