YFarmXIntelligence for an accelerating worldBriefings
YFarmX
YFarmX

AI Desk · Model Identity

NVIDIA: Nemotron 3 Ultra

nvidia/nemotron-3-ultra-550b-a55b · NVIDIATokenizer lineage measured
Tokenizer lineage (measured)
Mistral
Evidence confidence
High
Why high: an exact fingerprint shared with 21 models that declare the same family, without a second signal of a different kind yet.
Declared tokenizer tag
Other
Entered catalogue
4 June 2026
Last tested
3 September 2026
Measurement
Measured
Evidence
  • ✓ exact tokenizer signature shared with 21 models
Compare with Mistral: Mistral Medium 3.5

Download JSONReport a correction

Identity statement

NVIDIA: Nemotron 3 Ultra is NVIDIA's model. Its measured tokenizer sits with the Mistral group, which is evidence that it builds on the Mistral tokenizer. Tokenizer reuse is normal engineering, open vocabularies travel between labs, and this says nothing against NVIDIA's authorship of the model itself.

Evidence stack

Signals of different kinds, weighed together
SignalResultReference
Tokenizer signatureExact match with 21 other modelstk_0da8a999
API surfaceNo other entry shares this exact contractapi_4fe0bcd5
Context and output262,144 context · 182,520 max outputdeclared
Reasoning contractOptional · efforts high, medium · default highdeclared
Serving providers observedBaseTenmeasured 3 September 2026

Shares this fingerprint

Exact signature first; template-boundary shifts of the same signature beneath
Mistral: Codestral 2508mistralai/codestral-2508exact · 50/50Mistral: Devstral 2 2512mistralai/devstral-2512exact · 50/50Mistral: Ministral 3 14B 2512mistralai/ministral-14b-2512exact · 50/50Mistral: Ministral 3 3B 2512mistralai/ministral-3b-2512exact · 50/50Mistral: Ministral 3 8B 2512mistralai/ministral-8b-2512exact · 50/50Mistral Largemistralai/mistral-largeexact · 50/50Mistral Large 2407mistralai/mistral-large-2407exact · 50/50Mistral: Mistral Large 3 2512mistralai/mistral-large-2512exact · 50/50Mistral: Mistral Medium 3mistralai/mistral-medium-3exact · 50/50Mistral: Mistral Medium 3.5mistralai/mistral-medium-3-5exact · 50/50Mistral: Mistral Medium 3.1mistralai/mistral-medium-3.1exact · 50/50Mistral: Mistral Nemomistralai/mistral-nemoexact · 50/50Mistral: Sabamistralai/mistral-sabaexact · 50/50Mistral: Mistral Small 4mistralai/mistral-small-2603exact · 50/50Mistral: Mistral Small 3.1 24Bmistralai/mistral-small-3.1-24b-instructexact · 50/50Mistral: Mistral Small 3.2 24Bmistralai/mistral-small-3.2-24b-instructexact · 50/50Mistral: Mixtral 8x22B Instructmistralai/mixtral-8x22b-instructexact · 50/50Mistral: Voxtral Small 24B 2507mistralai/voxtral-small-24b-2507exact · 50/50NVIDIA: Nemotron 3 Nano 30B A3Bnvidia/nemotron-3-nano-30b-a3bexact · 50/50NVIDIA: Nemotron 3 Supernvidia/nemotron-3-super-120b-a12bexact · 50/50NVIDIA: Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightningexact · 50/50NVIDIA: Nemotron Nano 9B V2 (free)nvidia/nemotron-nano-9b-v2:freesame tokenizer · template shiftNVIDIA: Nemotron 3 Nano Omni (free)nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:freesame tokenizer · template shift

Closest measured models

Agreement across the mutually clean test strings
Mistral: Mistral Nemomistralai/mistral-nemo50/50 · 100%NVIDIA: Nemotron 3 Nano 30B A3Bnvidia/nemotron-3-nano-30b-a3b50/50 · 100%NVIDIA: Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightning50/50 · 100%NVIDIA: Nemotron 3 Supernvidia/nemotron-3-super-120b-a12b50/50 · 100%Mistral: Mistral Small 3.2 24Bmistralai/mistral-small-3.2-24b-instruct50/50 · 100%Mistral: Ministral 3 3B 2512mistralai/ministral-3b-251250/50 · 100%

Click a row to open the full comparison.

The measured fingerprint

Marginal prompt-token cost of each test string, grouped by script
All 50 rows

English and whitespace

en-prose14
en-long18
spaces-203
spaces-603
tabs-204
newlines-207
mixed-ws7

Digits

digits-99
digits-1212
digits-3030
digits-sep12
float-long22

CJK

zh-common16
zh-long20
zh-rare20
ja-kana11
ja-kanji13
ko13

Other scripts

ru12
ar9
he10
hi14
th17
el13

Emoji

emoji-basic20
emoji-skin40
emoji-zwj-family19
emoji-zwj-x357
emoji-flags32
emoji-prof35

Rare Unicode

math33
boxdraw30
combining14
cjk-ext-b16
surrogates53
zalgo36
rtl-mix6

Code

py-code25
py-indent15
json25
html16
regex49
camel6
snake7

Repetition and encodings

rare-word-x536
repeat-tok41
base6439
hex19
url19
uuid36

An x marks a row excluded as corrupt (caching or provider interference detected during measurement). Overhead subtracted: 17 prompt tokens.

Declared record

What the catalogue claims about this model

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

Modalities
text->text
Declared tokenizer
Other
Prompt price
$0.60 / M tokens
Completion price
$2.40 / M tokens

Supported parameters

frequency_penaltyinclude_reasoninglogit_biasmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_p

defaults: {"temperature":1,"top_p":0.95,"top_k":null,"frequency_penalty":null,"presence_penalty":null,"repetition_penalty":null}

Catalogue entryWeights on Hugging Face

History

Every observation, kept as taken
  • 4 June 2026Enters the OpenRouter catalogue with no declared family.
  • 3 September 2026Fingerprinted in the YFarmX catalogue sweep · 50 of 50 strings measured clean.
  • 3 September 2026YFarmX assessment: consistent with the Mistral family, high confidence.

Cite this page as the evidence record for NVIDIA: Nemotron 3 Ultra: the URL is stable, measurements are dated, and revisions append to the history above. Method: tokenizer fingerprinting, v1, 50 strings.