NVIDIA announced Nemotron 3 Ultra, its next open model, at the keynote.
Resolved true The record shows it happened or was so.
Factual: Statements about the present or the past, checked against the record: did they hold up?
Said for an organisation they lead or work for; it counts on both records.
What would count as true or false
Judged as of 2026-06-01. True: NVIDIA announced a model called Nemotron 3 Ultra at that keynote, on or about 1 June 2026 (shown by the keynote recording or transcript, or an NVIDIA press release or blog post dated that day). Mostly true: Nemotron 3 Ultra was announced at that event, but a minor detail differs, such as it being presented as an availability date, a release or a preview of a model announced earlier, or a slight variation in the name. Mixed: Huang talked about Nemotron 3 Ultra at the keynote, but it had already been fully announced earlier, so this was not a new announcement in substance. Mostly false: the model announced was a different Nemotron variant (e.g., Super or Nano), or the Ultra had been launched well before and was only mentioned. False: no Nemotron 3 Ultra was announced at or around the keynote.
Set before the research began.
Evidence (4)
-
E1 Enterprise Software Leaders Build AI Agents With NVIDIA NVIDIA (investor.nvidia.com, GlobeNewswire release) · 1 June 2026 Own record
“NVIDIA today unveiled new open models and datasets for always-on agents”
-
E2 Nvidia Nemotron 3 Ultra Tops US Open Models but Trails China Implicator.ai · 2 June 2026
“Nvidia unveiled Nemotron 3 Ultra at its GTC Taipei keynote on Monday and released a benchmark, run with the evaluation firm Artificial Analysis”
-
E3 Nvidia's best model is now live Ground News (aggregator) · 29 October 2026
“After pre-announcing Nemotron 3 Ultra, a 550-billion-parameter open-weight mixture-of-experts model, at Computex, Nvidia on Thursday released the model on platforms like Hugging Face, ModelScope, OpenRouter (with a free endpoint), and build.nvidia.com.”
-
E4 📅 ThursdAI - Jun 4 - NVIDIA drops Nemotron 3 Ultra (550B open), Microsoft becomes a frontier lab, Ideogram 4 goes open, Agent Arena & more ThursdAI (Alex Volkov newsletter) · 5 June 2026
“And then today, NVIDIA dropped Nemotron 3 Ultra, their latest 550B open weights model, data and training and Arena published a new agentic eval leaderboard and we got a new Gemma 4 12B.”
Reasoning
E2, an independent outlet writing on 2026-06-02, reports that Nvidia unveiled Nemotron 3 Ultra at its GTC Taipei keynote on Monday, which was 1 June 2026. E3 also says the model was pre-announced at Computex, and E4 says the weights were released later that week. NVIDIA's own release of 1 June (E1) says it unveiled new open models that day. None of the evidence shows that Nemotron 3 Ultra had been fully announced before the keynote, so the claim meets the criteria for true.
Cite this claim
Jensen Huang, “NVIDIA announced Nemotron 3 Ultra, its next open model, at the keynote.” (said 1 June 2026). Verdict: resolved true. Reliability API, claim c_a38b8cfe9ca4, method 0.1.0+03d23c0e, dataset release v1.1.2. https://epistacklabs.com/claims/c_a38b8cfe9ca4
This claim is in dataset release v1.1.2, with the same verdict.
Reply to this claim
More factual claims by Jensen Huang
-
Mostly true
“our architecture is fungible in the sense that it runs every AI model, runs all data processing and parallel processing tasks.”
NVIDIA's architecture runs every AI model and all data processing and parallel processing tasks.
-
Mixed
“A long time ago, NVIDIA used to be a GPU company.”
-
Resolved true
“NVIDIA Open Shell is open source.”
-
Resolved true
“It is the reason why we’re working with Cadence so hard to accelerate all of their tools on CUDA, because the agents are impatient.”
NVIDIA is working with Cadence to accelerate Cadence's tools on CUDA.
-
Mostly true
“Tax expense included represents impact from OBBBA (One Big Beautiful Bill Act).”
Verdicts are made automatically; how verdicts are made.