Up to date
nvidia NVIDIA Open Model License (nvidia-open-model-license) out 17 months ago
8.72B parameters, images and text in, text out.
Verified 2026-10-25 Rank #453 in LLMs · score 77.9 method v1 Sources 3 Downloads 30d 82.7k · #334 of 619 tracked
Picked by rule from 25 tracked pieces: the lab's own post, the most shared, a field report, the latest.
The 3 picks above, plus these 22. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 1 quantized builds and 16 Spaces.
Top Spaces running it: Gapeleon/Llama-3.1-Nemotron-Nano-VL-8B-V1-Demo ♥ 7 · ArtShumov/Whyx-PROmpTea ♥ 3 · ayushKishor/plutoV2_miniProject_3rd-yr ♥ 2 · shiva0013/YT-AI-Automation ♥ 1
Weights only, an estimate: leave room for the KV cache. Method
About 5.2 GB at INT4, 11 GB at FP8, 21 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2025-06-03), amber triangles are 10 dated articles (each links to it), grey ticks are repo commits.
3 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
| Benchmark | Value | Setting / notes | Reported by | Date |
|---|---|---|---|---|
| author-reported AI2D · Understand images and video | 85.0% percent | — | model author | 2026-10-07 |
| author-reported ChartQA · Understand images and video | 86.3% percent | — | model author | 2026-10-07 |
| author-reported DocVQA · Extract text from documents | 91.2% percent | — | model author | 2026-10-07 |
| author-reported InfoVQA · Extract text from documents | 77.4% percent | — | model author | 2026-10-07 |
| author-reported MMMU · Understand images and video | 48.2% percent | — | model author | 2026-10-07 |
| author-reported OCRBench · Extract text from documents | 839 | — | model author | 2026-10-07 |
| author-reported OCRBench · Extract text from documents | 60.1% percent | — | model author | 2026-10-07 |
| author-reported OCRBench · Extract text from documents | 37.9% percent | — | model author | 2026-10-07 |
| author-reported Video-MME · Understand images and video | 54.7% percent | Calculated with 1 tile per image | model author | 2026-10-07 |
The model performs image summarization and text-image analysis under the NVIDIA Open Model License.[1][2] It utilizes a Transformer architecture with a C-RADIOv2-H vision encoder and a Llama-3.1-8B-Instruct language encoder.[3][4]
Generated 2026-10-07 15:12 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Listed in LLMs
| Pipeline tag | image-text-to-text |
|---|---|
| Library | transformers |
| Params by dtype | BF16: 8,721,768,454 |
| Size band | M |
| MoE | no |
| License | NVIDIA Open Model License (nvidia-open-model-license) (other) |
| Hub snapshot | 2026-10-25 |
Weights-only memory for Llama-3.1-Nemotron-Nano-VL-8B-V1: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 21 GB · Single GPU (<24 GB) |
|---|---|
| FP8 (1 byte/param) | 11 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 5.2 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
quantized: 1adapter: 1finetune: 1
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| ekryski/Llama-3.1-Nemotron-Nano-VL-8B-V1-4bit | quantized | -4bit | 8.72B | M | apache-2.0 | ekryski | 1 | 624 |
| Nayana-cognitivelab/Llama_Nemotron_SectionOCR_SFT_En_Kn_15000 | adapter | — | 8.72B | M | other | Nayana-cognitivelab | 0 | 128 |
| runanywhere/nemotron_nano_vl_8b_HNPU | finetune | — | — | — | other | runanywhere | 0 | 0 |
| Date | Commit | SHA |
|---|---|---|
| 2025-12-04 | Update README | 437f4e2 |
| 2025-06-13 | Update README | 8dd8d15 |
| 2025-06-12 | Update README | fbd4f16 |
| 2025-06-12 | Update README | fea8968 |
| 2025-06-10 | Update README | 58eef79 |
| 2025-06-09 | Update model class name | 961cb9f |
| 2025-06-05 | Update README | a02869b |
| 2025-06-05 | Update README and examples | 1fbc10a |
| 2025-06-03 | Update paths | 67c7cea |
| 2025-06-03 | Initial commit | 1e8e8e6 |
| 2025-06-03 | initial commit | 6a5079e |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency.
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:40 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:12 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
