NVIDIA › Nemotron 3 Ultra › Pricing

Nemotron 3 Ultra pricing

Cheapest host: OpenRouter · checked Oct 26, 2026
Benchmarks and alternatives →
$0.60 / $2.40per 1M tokens in / out
$0.37per task at thinking?
#9on OpenRouter by tokens · 5.9T / week

Price sheetper 1M tokens

Input$0.60OpenRouter
Output$2.40
Context · max output1M · 66Ktokens
Free tieryesfree OpenRouter route
Secondary zero quote; a free lab API is not confirmed.

Where to buy4 providers · per 1M tokens

ProviderInputOutputContext
nvidialab$0$0—
DeepInfra$0.50$2.20—
BaseTen$0.60$2.40203K
Venice$0.63$3.13256K

In NVIDIA's line-upper 1M in / out

Nemotron 3.5 Content Safety$0.20 / $0.20−81%
NVIDIA Nemotron Nano 12B v2 VL$0.20 / $0.60−71%
Nemotron 3 Super 120B A12B$0.13 / $1.15−64%
Nemotron 3 Ultra$0.60 / $2.40this model
Llama 3.1 Nemotron 70B Instruct$1.20 / $1.20+14%
OpenCodeReasoning Nemotron 1.1 32B$2.72 / $4.803.1×
OpenReasoning Nemotron 32B$2.72 / $4.803.1×
<

Cheaper for the same workvs Nemotron 3 Ultra

GPT-6.1 Sol high-13%score ×2.2$0.32
MiMo v2.6 Pro thinkingOpen-64%score ×2$0.13
GPT-6.1 Terra xhigh-47%score +98%$0.19
DeepSeek V4.1 Flash 1017 maxOpen-29%score +97%$0.26
Claude Haiku 5.5 max-42%score +89%$0.21

Pricing questions

How much does Nemotron 3 Ultra cost?›

$0.60 per 1M input tokens and $2.40 per 1M output tokens. One Intelligence Index task costs about $0.37 at the default effort.

Is Nemotron 3 Ultra free to use?›

Yes, with limits: free OpenRouter route. Free route; provider availability and account rate limits apply.

What is a cheaper alternative to Nemotron 3 Ultra?›

GPT-6.1 Sol at high effort costs -13% per task with a score ×2.2 on Humanity's Last Exam.