# Live /models (2026-06-25): none|low|medium|high; no dedicated reasoning budget.
# https://api.venice.ai/api/v1/models?type=text (accessed 2026-06-25)

name = "DeepSeek V3.2"
description = "DeepSeek chat model for instruction following, coding, and analysis"
family = "deepseek"
release_date = "2025-12-04"
last_updated = "2026-06-11"
attachment = false
reasoning = true
tool_call = true
structured_output = true
open_weights = true

[[reasoning_options]]
type = "effort"
values = ["none", "low", "medium", "high"]

[cost]
input = 0.33
output = 0.48
cache_read = 0.16

[limit]
context = 160_000
output = 32_768

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