NVIDIA › Pricing
NVIDIA API pricing
per 1M tokens · list prices from the lab · checked Oct 26, 2026
$0.025 / $0.12cheapest · Nemotron 3.5 Lightning
$0.06 / $0.35top ranked · Nemotron 3.5 Super VL
−83%cache read vs input
24models with a price
All NVIDIA modelsclick a row for its full price sheet
| Model | Input | Output | Cache read | Context | Released ↓ | Status |
|---|---|---|---|---|---|---|
Nemotron 3.5 Super VL | h$0.06 | $0.35 | $0.01 | 1M | Oct 2026 | |
NVIDIA Nemotron 3.5 Super VL 120B A12B NVFP4 | h!$0.08 | $0.45 | — | 1M | Oct 2026 | |
Nemotron 3.5 Lightning | h$0.025 | $0.12 | — | 1M | Aug 2026 | |
Nemotron 3 Ultra | h!$0.60 | $2.40 | — | 1M | Jun 2026 | |
Nemotron 3.5 Content Safety | h!$0.20 | $0.20 | — | 128K | Jun 2026 | |
Nemotron 3 Nano Omni 30B A3B Reasoning | h!$0 | $0 | — | 256K | Apr 2026 | |
Nemotron 3 Super 120B A12B | h$0.13 | $1.15 | — | 1M | Mar 2026 | |
Nemotron 3 Super Older | h$0.085 | $0.40 | — | 1M | Mar 2026 | Older |
Nemotron Nano 3 30B | h!$0.06 | $0.24 | — | 256K | Dec 2025 | |
Nemotron 3 Nano 30B A3B | h!$0.06 | $0.24 | — | 131K | Dec 2025 | |
NVIDIA Nemotron Nano 12B v2 VL | h$0.20 | $0.60 | — | 128K | Oct 2025 | |
NVIDIA Nemotron Nano 9B V2 | h!$0.04 | $0.16 | — | 131K | Aug 2025 | |
Llama Nemotron Super 49B v1.5 Older | h$0.40 | $0.40 | — | 128K | Jul 2025 | Older |
OpenReasoning Nemotron 32B | h!$2.72 | $4.80 | — | 131K | Jul 2025 | |
OpenCodeReasoning Nemotron 1.1 32B | h!$2.72 | $4.80 | — | 66K | Jun 2025 | |
Llama 3.1 Nemotron 70B Instruct | h$1.20 | $1.20 | — | 131K | Oct 2024 | |
Nemotron 4 340B Instruct | h$4.20 | $4.20 | — | 4K | Jun 2024 |
| Model | Type ↑ | Price | Released | Status |
|---|---|---|---|---|
Nemotron 3 Embed 1B BF16 | Embeddings · Embeddings | $0.015 / 1M tok | Jul 2026 | |
Llama Nemotron Embed VL 1B v2 | Embeddings · Multimodal | $0.01 / 1M tok | Jan 2026 | |
Nemotron 3.5 Embed 1B BF16 | Embeddings · Multimodal | $0.015 / 1M tok | Oct 2026 | |
Nemotron 3.5 Embed 8B BF16 | Embeddings · Multimodal | $0.035 / 1M tok | Oct 2026 | |
Canary Qwen 2.5B, NVIDIA | Speech · Speech to text | $0.001 / min | Jul 2025 | |
Parakeet TDT 0.6B V3, Togetherai | Speech · Speech to text | $0.002 / min | Aug 2025 | |
Parakeet RNNT 1.1B | Speech · Speech to text | $0.002 / min | Dec 2023 |
36 more without a listed price
NVIDIA pricing questions
How much does the NVIDIA API cost?›
NVIDIA API prices per 1M input / output tokens: Nemotron 3.5 Super VL $0.06 / $0.35, NVIDIA Nemotron 3.5 Super VL 120B A12B NVFP4 $0.08 / $0.45, Nemotron 3.5 Lightning $0.025 / $0.12, Nemotron 3 Ultra $0.60 / $2.40, Nemotron 3.5 Content Safety $0.20 / $0.20. Cached input is about 83% cheaper.
What is the cheapest NVIDIA model?›
Nemotron 3.5 Lightning at $0.025 / $0.12 per 1M tokens.
What is the most expensive NVIDIA model?›
OpenCodeReasoning Nemotron 1.1 32B at $2.72 / $4.80 per 1M tokens.
Is there a free NVIDIA API?›
Yes, for Nemotron 3 Ultra, Nemotron 3.5 Lightning, Nemotron 3.5 Super VL, Nemotron 3 Super: free OpenRouter route.
How much does Nemotron 3.5 Super VL cost?›
$0.06 / $0.35 per 1M input / output tokens, cache read $0.01.