NVIDIA · open weights
Nemotron 3 Super API pricing by provider
Nemotron 3 Super is available from 2 providers. The cheapest is DeepInfra at $0.085 input / $0.40 output per 1M tokens (BF16). The priciest host, DekaLLM, charges 1.1x the cheapest.
Prices from the OpenRouter API, retrieved September 15, 2026. Weights: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8.
| Provider | Precision | Input / 1M | Output / 1M | Document Q&A / month | Context | Max output |
|---|---|---|---|---|---|---|
| DeepInfra | bf16 | $0.085 | $0.40 | $46.00 | 262K | 16K |
| DekaLLM | fp8 | $0.08 | $0.45 | $45.50 | 262K | 236K |
Sorted by blended price (3 input : 1 output). Document Q&A assumes 50,000 questions a month with 8,000 input and 600 output tokens each. Precision is reported by the provider; “—” means not disclosed.
Frequently asked questions
- What is the cheapest Nemotron 3 Super API provider?
- DeepInfra is the cheapest healthy provider we track at $0.085 per 1M input tokens and $0.40 per 1M output tokens.
- How much does Nemotron 3 Super cost?
- Across 2 providers, input prices range from $0.08 to $0.085 per 1M tokens and output prices from $0.40 to $0.45.
- What is Nemotron 3 Super’s context window?
- Nemotron 3 Super supports up to 262K tokens, but some providers serve a smaller context window — check the table before choosing.
Other open-weight models
- DeepSeek V4.1 Flashfrom $0.15 / $0.60
- DeepSeek V4 Pro 0813from $0.96 / $2.88
- Kimi K3from $2.10 / $10.95
- GLM 5.3from $0.91 / $2.86
- GLM 5.3 Flashfrom $0.075 / $0.25
- Qwen3.8 27Bfrom $0.15 / $2.00
- MiniMax M3from $0.23 / $0.96
- gpt-oss-120bfrom $0.03 / $0.17
- gpt-oss-20bfrom $0.02 / $0.10
- Llama 4 Maverickfrom $0.1875 / $0.6525
- Llama 3.3 70B Instructfrom $0.10 / $0.32
- Gemma 4 31Bfrom $0.09 / $0.34
- Mistral Small 4from $0.15 / $0.60