# Nebius

HQ NL · serves 27 tracked open-weights models, 13 confirmed by its own catalog. · privacy policy

models
27
verified
13
of overall top 10
7
cheapest on
3 models
vs best $
1.25×
ctx (median)
125k
uptime
91.3%

# models

model rank $/M in $/M out $/M cache src
kimi-k3 #1 nebius
deepseek-v4-pro #2 nebius
glm-5.2 #3 nebius
kimi-k2.6 #4 nebius
minimax-m3 #6 nebius
kimi-k2.7-code #7 nebius
glm-5.1 #10 $1.40 $4.40 openrouter
minimax-m2.5 #11 $0.30 $1.20 $0.30 requesty
qwen3.5-397b-a17b #17 nebius
deepseek-r1 #33 $0.80 $2.40 litellm
gpt-oss-120b #35 $0.15 $0.60 openrouter
qwen3-235b-a22b-instruct-2507 #39 $0.20 $0.60 openrouter
deepseek-v3 #42 $0.50 $1.50 litellm
qwq-32b #48 $0.15 $0.45 litellm
gemma-3-27b-it #54 $0.10 $0.30 openrouter
deepseek-r1-0528 $0.80 $2.40 litellm
deepseek-v4-flash-0731 nebius
llama-3.1-405b-instruct $1.00 $3.00 litellm
llama-3.3-70b-instruct $0.13 $0.40 openrouter
nemotron-3-nano-30b-a3b $0.06 $0.24 openrouter
qwen2.5-72b-instruct $0.13 $0.40 litellm
qwen2.5-vl-72b-instruct $0.13 $0.40 litellm
qwen3-14b $0.08 $0.24 litellm
qwen3-235b-a22b $0.20 $0.60 litellm
qwen3-30b-a3b $0.10 $0.30 litellm
qwen3-32b $0.10 $0.30 $0.10 requesty
qwen3-next-80b-a3b-thinking $0.15 $1.20 $0.15 requesty

# recent changes

when (UTC)modelchange
Sep 3 16:33 qwen2.5-vl-72b-instruct $/M input $0.25 → $0.13 ↓ 48.0% [source switch]
Sep 3 16:33 qwen2.5-vl-72b-instruct $/M output $0.75 → $0.40 ↓ 46.7% [source switch]
Sep 3 09:55 deepseek-v3.2 no longer listed
Sep 3 09:55 deepseek-v4-flash no longer listed
Sep 3 09:55 glm-5 no longer listed
Sep 3 09:55 kimi-k2.5 no longer listed

all providers · machine-readable: providers.json