Nemotron 3 Ultra nemotron-3-ultra-550b-a55b

nvidia · nvidia

unranked — no quality signals yet

params
560.5B
arch
moe
context
256k
license
other open weights
released
Jun 2026
downloads/30d
1.2M

# architecture

attn shared GQA 64q/2kv router top-22 of 512 +1 shared expert ×512 expert out ctx 262,144 560.5B pool
d_model
8192
heads
64q / 2kv
head dim
128
experts
top-22 of 512 + 1 shared
vocab
131072
family
nemotron_h

signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.

# why ranked

Overall: provisional — score 44.0 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_intelligence 0.70 —
bench_composite 0.30 44.0

benchmark panel evidence:

complete panelscore (0–100)
browsecomp-v144.0

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

OGM score · last 45 days
OGM score over 45 days

# benchmarks

benchmarkscoresourcedate
Aa-lcr 0.7 / 1 LLM Stats —
Apex 0.8 / 1 LLM Stats —
Browsecomp 0.4 / 1 LLM Stats —

# where to run

providerquantctx$/M in$/M out$/M cacheprice srctpsuptime✓
Nebius — ——— nebius — — ✓
Nvidia 1024k ——— via requesty — —
DeepInfrafp4 256k $0.40$1.76$0.08 via openrouter — 100.0%
DeepInfrafp4 256k $0.50$2.20$0.10 via openrouter — 99.9%
BaseTenfp4 202k $0.60$2.40$0.12 via openrouter — 99.8%
BaseTenfp4 202k $0.60$2.40$0.12 via openrouter — 99.9%
Venicefp8 250k $0.62$3.12$0.19 via openrouter — 99.7%
Together 512k $0.60$3.60$0.20 via modelsdev — —

sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price

source aliases
llmstats
nemotron-3-ultra-550b-a55b
openrouter
nvidia/nemotron-3-ultra-550b-a55b