NVIDIA API Pricing
Real-time API token pricing for NVIDIA AI models including Llama Nemotron and other NVIDIA-hosted models. Compare input/output costs via OpenRouter.
13 active models · Data updated hourly
About NVIDIA
NVIDIA hosts and fine-tunes AI models through NVIDIA NIM (NVIDIA Inference Microservices) and makes them available via API. Their offerings include fine-tuned variants of Meta Llama and other open-weight models optimized for NVIDIA hardware. NVIDIA also produces their own models like Nemotron, designed for enterprise workloads on NVIDIA GPU infrastructure.
| Model | Released | Context | Input $/1M | Output $/1M | Modalities |
|---|---|---|---|---|---|
| NVIDIA: Nemotron 3.5 Super VL | Oct 2026 | 1.0M | $0.150 | $0.500 | Vision, Open Source |
| NVIDIA: Switchyard | Sep 2026 | 1M | <$0.01 | <$0.01 | Text |
| NVIDIA: Nemotron 3.5 Lightning | Aug 2026 | 262K | $0.070 | $0.200 | Open Source |
| NVIDIA: Nemotron 3.5 Content Safety | Jun 2026 | 131K | $0.200 | $0.200 | Vision, Open Source |
| NVIDIA: Nemotron 3 Ultra | Jun 2026 | 262K | $0.600 | $2.40 | Open Source |
| NVIDIA: Nemotron 3 Super | Mar 2026 | 262K | $0.300 | $2.40 | Open Source |
| NVIDIA: Nemotron 3 Nano 30B A3B | Dec 2025 | 262K | $0.060 | $0.240 | Open Source |
| NVIDIA: Nemotron 3.5 Super VL (free) | Oct 2026 | 1M | Free | Free | Vision, Free, Open Source |
| NVIDIA: Nemotron 3.5 Lightning (free) | Aug 2026 | 1M | Free | Free | Free, Open Source |
| NVIDIA: Nemotron 3.5 Content Safety (free) | Jun 2026 | 128K | Free | Free | Vision, Free, Open Source |
| NVIDIA: Nemotron 3 Ultra (free) | Jun 2026 | 1M | Free | Free | Free, Open Source |
| NVIDIA: Nemotron 3 Nano Omni (free) | Apr 2026 | 256K | Free | Free | Vision, Audio, Free, Open Source |
| NVIDIA: Nemotron 3 Super (free) | Mar 2026 | 262K | Free | Free | Free, Open Source |