# Tencent

HQ CN · datacenters SG · serves 14 tracked open-weights models. status page · privacy policy

models
14
verified
—
of overall top 10
2
cheapest on
4 models
vs best $
2.48×
ctx (median)
1000k
uptime
99.9%

# models

model rank $/M in $/M out $/M cache src ✓
deepseek-v4-pro #1 $1.74 $3.48 — litellm —
minimax-m3 #2 $0.30 $1.20 — litellm —
deepseek-v4-flash — $0.14 $0.28 — litellm —
deepseek-v4-pro-0813 — $1.32 $3.96 — litellm —
deepseek-v4.1-flash — $0.30 $1.20 — litellm —
glm-5.2 — $1.40 $4.40 — litellm —
glm-5.3 — $1.40 $4.40 — litellm —
glm-5.3-flash — $0.15 $0.50 — litellm —
hy-mt2-1.8b — $0.04 $0.18 — openrouter —
hy-mt2-30b-a3b — $0.07 $0.29 — openrouter —
hy-mt2-7b — $0.07 $0.29 — openrouter —
hy4-preview — $0.75 $2.25 $0.04 openrouter —
kimi-k2.7-code — $0.95 $4.00 — litellm —
kimi-k3 — $3.00 $15.00 — litellm —

# recent changes

when (UTC)modelchange
Nov 1 17:59 hy4-preview $/M input $0.83 → $0.75 ↓ 10.0%
Nov 1 17:59 hy4-preview $/M output $2.50 → $2.25 ↓ 10.0%
Nov 1 06:26 hy4-preview $/M input $0.75 → $0.83 ↑ 11.1%
Nov 1 06:26 hy4-preview $/M output $2.25 → $2.50 ↑ 11.1%
Oct 31 21:45 hy4-preview $/M input $0.83 → $0.75 ↓ 10.0%
Oct 31 21:45 hy4-preview $/M output $2.50 → $2.25 ↓ 10.0%
Oct 31 05:58 hy4-preview $/M input $0.75 → $0.83 ↑ 11.1%
Oct 31 05:58 hy4-preview $/M output $2.25 → $2.50 ↑ 11.1%
Oct 30 22:30 hy4-preview $/M input $0.83 → $0.75 ↓ 10.0%
Oct 30 22:30 hy4-preview $/M output $2.50 → $2.25 ↓ 10.0%

all providers · machine-readable: providers.json