← All benchmarks
Nemotron 3 Nano Omni 30B A3B Reasoning
NVIDIA · Released 2026-04-29 · Independent benchmark scores from Artificial Analysis
Intelligence Index
10.3
Output speed
239.1
tokens/sec
Blended price
$0.45
per 1M tokens
Time to first token
0.40s
Model metadata
| Model ID | 8c748e53-61ae-48b8-af8d-eb8298b1e9db |
|---|---|
| Slug | nemotron-3-nano-omni-30b-a3b |
| Release date | 2026-04-29 |
| Provider | NVIDIA |
| Provider slug | nvidia |
| Output speed | 239.1 tok/s |
| Time to first token | 0.40s |
| Time to first answer | 8.77s |
| Blended price ($/1M) | $0.45 |
| Input price ($/1M) | $0.30 |
| Output price ($/1M) | $0.90 |
All evaluation scores (18)
| aime | — |
|---|---|
| aime 25 | — |
| artificial analysis coding index | 13.80 |
| artificial analysis intelligence index | 10.30 |
| artificial analysis math index | — |
| gpqa | 46.9% |
| hle | 4.8% |
| ifbench | 63.2% |
| lcr | 39.7% |
| livecodebench | — |
| math 500 | — |
| mmlu pro | — |
| scicode | — |
| tau banking | — |
| tau2 | 45.3% |
| terminalbench hard | 8.3% |
| terminalbench v2 1 | 6.7% |
| terminalbench v4 0 | — |
Source: artificialanalysis.ai/models/nemotron-3-nano-omni-30b-a3b
Benchmark data from Artificial Analysis.