Providers
Create a free accountBrowse all models
NVIDIA API
NVIDIA’s Nemotron models through one unified API key. Open, capable models for reasoning and agents at pay-as-you-go prices.
5 modelsOpenAI- & Anthropic-compatiblePay-as-you-go
| Model | Tier | Context | Price in / out (1M) | |
|---|---|---|---|---|
| Nemotron 3 Nano 30B A3B | Paid | 262K | $0.07 / $0.27 | View |
| Nemotron 3 Ultra 550B A55B | Paid | — | $0.66 / $2.64 | View |
| Nemotron 3.5 Content Safety | Paid | — | $0.22 / $0.22 | View |
| Nemotron 3.5 Lightning | Paid | — | $0.08 / $0.22 | View |
| Nemotron 3.5 Super VL | Paid | — | $0.17 / $0.55 | View |
Call NVIDIA in three lines
Same code as the OpenAI SDK — change the base URL and key, pick a model id:
from openai import OpenAI
client = OpenAI(
base_url="https://api.airforce/v1",
api_key="YOUR_KEY",
)
resp = client.chat.completions.create(
model="nemotron-3-nano-30b-a3b",
messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)Frequently asked questions
- How do I use NVIDIA models through Api.Airforce?
- Point any OpenAI-compatible client at https://api.airforce/v1 and set the model id (for example "nemotron-3-nano-30b-a3b"). Your existing OpenAI SDK works unchanged — you only swap the base URL and API key. The same key also speaks the Anthropic Messages API.
- Are NVIDIA models free?
- NVIDIA models are premium — you add credits (a monthly plan or a pay-as-you-go top-up) to call them. A free account can still use the open free-tier models from other families at $0.
- How much do NVIDIA models cost on Api.Airforce?
- Billing is pay-as-you-go per token, at rates that sit well below the official APIs. The exact per-model input and output prices are listed in the table above and update automatically as upstream prices change.
- Which NVIDIA models are available?
- Currently 5 NVIDIA models, including Nemotron 3 Nano 30B A3B, Nemotron 3 Ultra 550B A55B, Nemotron 3.5 Content Safety. The catalog is refreshed as new models launch, so the list above is always current.