This model is deprecated. We still benchmark its performance, but less often: every 18 hours instead of every 6.
NVIDIA has launched a newer release, Nemotron 3.5 Super VL. We suggest considering it instead.
Nemotron 3 Super 120B A12B: Intelligence, Performance & Price Analysis
Model summary
Verbosity
Nemotron 3 Super 120B A12B (Reasoning) supports text input, outputs text, and has a 1M tokens context window.
Pricing for Nemotron 3 Super 120B A12B (Reasoning) is $0.30 per 1M input tokens (moderately priced, median: $0.30) and $0.90 per 1M output tokens (somewhat expensive, median: $0.80).
| Reasoning | Yes |
|---|---|
| Input modality | Supports: text |
| Output modality | Supports: text |
| Context window | 1M ~1500 A4 pages of size 12 Arial font |
| Total parameters | 120.6B |
| Active parameters | 12.7B Number of parameters active per token during inference |
| License | Nvidia Nemotron Open Model License |
| Model weights | Hugging Face |
Metrics are compared against models of the same class:
- Non-reasoning models → compared only with other non-reasoning models
- Reasoning models → compared across both reasoning and non-reasoning
- Open weights models → compared only with other open weights models of the same size class:
- Tiny: ≤4B parameters
- Small: 4B–40B parameters
- Medium: 40B–150B parameters
- Large: >150B parameters
- Proprietary models → compared across proprietary and open weights models of the same price range, using a blended 3:1 input/output price ratio:
- <$0.15 per 1M tokens
- $0.15–$1 per 1M tokens
- >$1 per 1M tokens
Intelligence
Artificial Analysis Intelligence Index
Artificial Analysis Intelligence Index by Open Weights / Proprietary
Intelligence Index vs. Cost per Task
Evaluations
Agentic knowledge work, Elo
Agentic SaaS workflows
Reasoning & knowledge
Professional document reasoning, All-pass
Agentic coding & terminal use
Agentic business operations
Agentic real-world work tasks, Elo
Kubernetes incident root-cause analysis
Openness Index
Artificial Analysis Openness Index: Score
Cost
Cost per Intelligence Index Task
Cost to Run Artificial Analysis Intelligence Index
Pricing: Cache Hit, Input, and Output
Performance
Measured by time per task (minutes), output speed (tokens per second), time to first answer token and time to first token (seconds)
Time per Intelligence Index Task
Output Speed
Time to First Answer Token
Context Window
Context Window
Model Size (Open Weights Models Only)
Model Size: Total and Active Parameters
Further details
Weights | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Nemotron 3 Super 120B A12B (Reasoning) | - | 120.6B 12.7B active at inference time | 1M | $0.20 | - | 178 |
Frequently Asked Questions
Common questions about Nemotron 3 Super 120B A12B
Nemotron 3 Super 120B A12B was released on March 11, 2026.
Nemotron 3 Super 120B A12B was created by NVIDIA.
Nemotron 3 Super 120B A12B has a time to first answer token of 9.22s (based on the median across providers serving the model), which is somewhat higher than average compared to other open weight models of similar size (median: 8.18s).
Nemotron 3 Super 120B A12B costs $0.30 per 1M input tokens (better than average, median: $0.30) and $0.90 per 1M output tokens (better than average, median: $0.80), based on the median across providers serving the model.
Nemotron 3 Super 120B A12B costs $0.30 per 1M input tokens and $0.90 per 1M output tokens (based on the median across providers serving the model). For a blended rate (7:2:1 cache hit/input/output ratio), this is $0.20 per 1M tokens. Pricing may vary by provider. Compare provider pricing
Yes, Nemotron 3 Super 120B A12B is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
Nemotron 3 Super 120B A12B supports text input.
Nemotron 3 Super 120B A12B supports text output.
No, Nemotron 3 Super 120B A12B does not support image input. It can only process text.
No, Nemotron 3 Super 120B A12B is not multimodal. It only supports text input.
Nemotron 3 Super 120B A12B has a context window of 1M tokens. This determines how much text and conversation history the model can process in a single request.
Yes, Nemotron 3 Super 120B A12B is open weights. The model weights are publicly available and can be downloaded for self-hosting.
Nemotron 3 Super 120B A12B has 120.6B parameters (12.7B active).
Nemotron 3 Super 120B A12B is a Mixture of Experts (MoE) model with 120.6B total parameters, but only 12.7B active parameters are used during inference.
Nemotron 3 Super 120B A12B is released under the Nvidia Nemotron Open Model License license. This license allows commercial use. View license
Yes, Nemotron 3 Super 120B A12B is available via API through 3 providers. Compare API providers
Nemotron 3 Super 120B A12B is available through 3 API providers. Compare providers