MDL
NVIDIA Nemotron 3 Super 120B A12B
NVIDIA
Verified against Artificial Analysis · Jul 13, 2026
Capability
76
percentile index
Price /M
$0.25 / $0.775
in / out
Context
—
Avg score
51.9
7 benchmarks
AA Intelligence
25.4
index
Speed
147 tok/s
TTFT 1s
Released
Mar 11, 2026
About
NVIDIA Nemotron 3 Super 120B A12B is a open-weights model in the NVIDIA family from NVIDIA. Benchmarked on 7 evals, averaging 51.9.
Benchmark scores · 7
avg 51.9 — higher is betterThe research behind this model
all papers mentioning it →Tokenizer Fertility and Zero-Shot Performance of Foundation Models on Ukrainian Legal Text: A Comparative StudyMay 14, 2026HEBATRON: A Hebrew-Specialized Open-Weight Mixture-of-Experts Language ModelMay 11, 2026Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget ControlMay 8, 2026Beyond the Black Box: Interpretability of Agentic AI Tool UseMay 7, 2026NVIDIA Nemotron 3: Efficient and Open IntelligenceDec 24, 2025
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
Performance & price
Output speed147 tok/s
Latency (TTFT)1s
First answer15s
Price in / out$0.25 / $0.775 per 1M
Blended price$0.381 / 1M
Coding index37.7
Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →
History
full ledger →Capability over time
Append-only ledger — every observed change to this model's numbers.
API providers
Artificial Analysis$0.775/M