# Sources (accessed 2026-08-16):
#   https://docs.mistral.ai/models/model-cards/voxtral-small-25-07
#   https://mistral.ai/news/voxtral/
# Field values mirror Mistral's own first-party host entry in this repo
# (providers/mistral/models/voxtral-small-latest.toml); host-scoped keys
# (cost, status) are intentionally left to the provider files.
name = "Voxtral Small (latest)"
description = "Instruct model with native audio input for speech understanding and tool use"
family = "voxtral"
release_date = "2025-07-15"
last_updated = "2025-07-15"
attachment = true
reasoning = false
temperature = true
tool_call = true
open_weights = true

[limit]
context = 32_000
output = 32_000

[modalities]
input = ["text", "audio"]
output = ["text"]