+1 (415) 347-6981Get started
Fixed prices · You own the code · Proposal in 24h hello@startrise.io →
AI Model Release Tracker US & Canada · updated
  • 13Releases tracked
  • Nov 12Latest · Nemotron-Labs-Sigma (3B and 8B)
  • 100%Open-weight
  • USSanta Clara, United States

NVIDIA is a Santa Clara, United States-based computing company, founded in 1993, that develops the Nemotron family of open AI models alongside its chips, systems and software. Nemotron spans the Nemotron 3 Nano, Super and Ultra reasoning models, the Nemotron 3.5 Lightning, Super VL and Bolt models for agents and vision, and Nemotron Retriever models for embedding, reranking and document parsing. NVIDIA also releases Kumo foundation models for tabular and relational enterprise data. Every tracked NVIDIA release is open-weight. Many are published on its Hugging Face organization and offered through NVIDIA NIM.

NVIDIA (Nemotron)

United States US & Canada research.nvidia.com/labs/nemotron

13 releases latest Nov 12, 2026

  1. Nemotron-Labs-Sigma (3B and 8B)

    Research continuous diffusion language models that denoise text in a 16-D embedding space and decode in parallel blocks; 14 checkpoints (base, instruct, distilled). Nemotron Open Model License.

    • LLM
    • Open weights
    • 4K ctx
  2. Cosmos 3 Nano (Transfer-Auto, Sim-Auto, Sim-Transfer and Sim-Bimanual)

    Four 16B open-weight Cosmos 3 Nano checkpoints for interactive world simulation (multi-view driving, LiDAR, simulator-to-RGB, bimanual robot); OpenMDW1.1. HF collection also dated Oct 20.

    • Video
    • Open weights
  3. Nemotron 3.5 Transcribe

    Promptable open speech model (30B MoE, ~4B active): transcription in 40 languages, translation, speaker labels and audio Q&A; first among open models on Open ASR (English).

    • Audio
    • Open weights
  4. Nemotron 3.5 Bolt E4B 4GB

    Open small model for local conversational agents, built on Gemma 4 E4B with Nemotron post-training; runs in about 4GB (Q4_K_M), 35 languages, configurable reasoning.

    • LLM
    • Open weights
    • 64K ctx
  5. Nemotron 3.5 Embed (8B and 1B), Rerank 8B and Parse

    Open Nemotron Retriever models for multimodal (text and image) embedding, reranking and 631M document parsing; Rerank cuts agent token cost about 14%. On Hugging Face, NIM and OpenRouter.

    • Embedding
    • Open weights
    • API
  6. Kumo Relational

    Pre-trained relational foundation model that predicts churn, demand or fraud on enterprise databases in-context, no training. Beats supervised baselines on 41 tasks; on Hugging Face and NIM.

    • Other
    • Open weights
    • API
  7. Nemotron Labs Dory1.0-4B

    Experimental looped (latent-recurrent) 4B model: a shared block is re-applied to trade inference compute for accuracy from one checkpoint. OpenMDW-1.1 license; BF16, MXFP8, NVFP4 and GGUF.

    • LLM
    • Open weights
    • 262K ctx
  8. Nemotron 3.5 Super VL and Nemotron 3.5 Bolt 74B-A9B 24GB

    Super VL: open 120B-A12B vision-language model, 1M context. Bolt: 74B-A9B local agent model in 24GB via native 2-bit. Bolt E4B 4GB due Oct 23.

    • Multimodal
    • Open weights
  9. Kumo Tabular (Small, Medium and Large)

    Open tabular foundation models (28M–215M) for in-context classification/regression; OpenMDW-1.1; #1 on TabArena (ELO 1950).

    • Other
    • Open weights
  10. Nemotron 3.5 Lightning (30B-A3B)

    Open 30B MoE with 3B active, built for high-volume low-latency agents; MTP speculative decoding, NVFP4 and BF16 checkpoints.

    • LLM
    • Open weights
  11. Nemotron 3 Ultra

    • Reasoning
    • Open weights
  12. Nemotron 3 Super

    • Reasoning
    • Open weights
  13. Nemotron 3 Nano

    • Reasoning
    • Open weights
    • 1M ctx

NVIDIA (Nemotron) FAQ

What is NVIDIA (Nemotron)'s latest AI model?
NVIDIA (Nemotron)'s most recent release in the Startrise tracker is Nemotron-Labs-Sigma (3B and 8B), released Nov 12, 2026.
How many AI models has NVIDIA (Nemotron) released?
Startrise tracks 13 notable NVIDIA (Nemotron) releases since January 2025, from Nemotron 3 Nano (Dec 15, 2025) to Nemotron-Labs-Sigma (3B and 8B).
Does NVIDIA (Nemotron) release open-weight models?
Yes. 13 of 13 tracked NVIDIA (Nemotron) releases (100%) ship with downloadable weights; the most recent is Nemotron-Labs-Sigma (3B and 8B).
Where is NVIDIA (Nemotron) based?
NVIDIA (Nemotron) is based in Santa Clara, United States.