Their blog goes into detail but it shows up to 4x throughput over DeepSeek-R1 with better benchmarks.
Blog:\nhttps://developer.nvidia.com/blog/build-enterprise-ai-agents-with-advanced-open-nvidia-llama-nemotron-reasoning-models/
The model is available on HF and as a NIM. Has anyone tried it?
HF:\nhttps://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1
NIM:\nhttps://build.nvidia.com/nvidia/llama-3_1-nemotron-ultra-253b-v1"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia Just Released Llama Nemotron Ultra"}},"_tags":["story","author_devaniranjan","story_43627318","ask_hn"],"author":"devaniranjan","children":[43627325],"created_at":"2025-04-08T23:02:40Z","created_at_i":1744153360,"num_comments":1,"objectID":"43627318","points":13,"story_id":43627318,"story_text":"NVIDIA just released Llama 3.1 Nemotron Ultra (253B parameter model) that\u2019s showing great performance on GPQA-Diamond, AIME, and LiveCodeBench.
Their blog goes into detail but it shows up to 4x throughput over DeepSeek-R1 with better benchmarks.
Blog:\nhttps://developer.nvidia.com/blog/build-enterprise-ai-agents-with-advanced-open-nvidia-llama-nemotron-reasoning-models/
The model is available on HF and as a NIM. Has anyone tried it?
HF:\nhttps://huggingface.co/nvidia/Llama-3_1-Nemotron-Ultra-253B-v1
NIM:\nhttps://build.nvidia.com/nvidia/llama-3_1-nemotron-ultra-253b-v1","title":"Nvidia Just Released Llama Nemotron Ultra","updated_at":"2025-04-10T19:25:51Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hochmartinez"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia's new Llama-3.1 Nemotron Ultra outperforms DeepSeek R1 at half the size"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://venturebeat.com/ai/nvidias-new-llama-3-1-nemotron-ultra-outperforms-deepseek-r1-at-half-the-size/"}},"_tags":["story","author_hochmartinez","story_43653944"],"author":"hochmartinez","children":[43659051,43662392],"created_at":"2025-04-11T14:06:25Z","created_at_i":1744380385,"num_comments":2,"objectID":"43653944","points":11,"story_id":43653944,"title":"Nvidia's new Llama-3.1 Nemotron Ultra outperforms DeepSeek R1 at half the size","updated_at":"2025-04-12T11:30:28Z","url":"https://venturebeat.com/ai/nvidias-new-llama-3-1-nemotron-ultra-outperforms-deepseek-r1-at-half-the-size/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia Nemotron 3-Nano 30B-A3B-BF16"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16"}},"_tags":["story","author_tosh","story_46839668"],"author":"tosh","children":[46841206],"created_at":"2026-01-31T19:04:45Z","created_at_i":1769886285,"num_comments":1,"objectID":"46839668","points":10,"story_id":46839668,"title":"Nvidia Nemotron 3-Nano 30B-A3B-BF16","updated_at":"2026-03-05T23:28:07Z","url":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rvnx"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia releases weights for Llama-3.1-Nemotron-70B-Instruct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct"}},"_tags":["story","author_rvnx","story_41862271"],"author":"rvnx","children":[41862291,41862541,41868871],"created_at":"2024-10-16T18:38:06Z","created_at_i":1729103886,"num_comments":3,"objectID":"41862271","points":9,"story_id":41862271,"title":"Nvidia releases weights for Llama-3.1-Nemotron-70B-Instruct","updated_at":"2024-10-18T16:05:31Z","url":"https://huggingface.co/nvidia/Llama-3.1-Nemotron-70B-Instruct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"andsoitis"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia Nemotron 3 Ultra \u2013 open weight model"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://research.nvidia.com/labs/nemotron/Nemotron-3-Ultra/"}},"_tags":["story","author_andsoitis","story_49034772"],"author":"andsoitis","children":[49035318,49045447],"created_at":"2026-07-24T12:44:17Z","created_at_i":1784897057,"num_comments":1,"objectID":"49034772","points":9,"story_id":49034772,"title":"Nvidia Nemotron 3 Ultra \u2013 open weight model","updated_at":"2026-07-25T07:51:54Z","url":"https://research.nvidia.com/labs/nemotron/Nemotron-3-Ultra/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"KenHV"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia Llama-3.1-Nemotron-70B-Instruct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://build.nvidia.com/nvidia/llama-3_1-nemotron-70b-instruct"}},"_tags":["story","author_KenHV","story_41860741"],"author":"KenHV","children":[41860753],"created_at":"2024-10-16T16:05:14Z","created_at_i":1729094714,"num_comments":0,"objectID":"41860741","points":9,"story_id":41860741,"title":"Nvidia Llama-3.1-Nemotron-70B-Instruct","updated_at":"2024-10-17T13:17:41Z","url":"https://build.nvidia.com/nvidia/llama-3_1-nemotron-70b-instruct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"alexwatson405"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"NVIDIA just open sourced NeMo Data Designer, the synthetic data framework used internally to build both pre-training and post-training datasets for Nemotron.
It lets you define an entire synthetic data pipeline directly in Python: structured outputs, statistical samplers, LLM-generated columns, dependency-aware field relationships, Python/SQL/remote validators, and optional LLM-as-judge scoring. Supports quick preview mode for fast iteration before scaling up.
Install:
```\npip install data-designer\n```
A minimal example:
```\nfrom data_designer.essentials import *
data_designer = DataDesigner()\nconfig = DataDesignerConfigBuilder()
config.add_column(\n SamplerColumnConfig(\n name="product_category",\n sampler_type=SamplerType.CATEGORY,\n params=CategorySamplerParams(\n values=["Electronics", "Clothing", "Home & Kitchen", "Books"]\n ),\n )\n)
config.add_column(\n LLMTextColumnConfig(\n name="review",\n model_alias="nvidia-text",\n prompt="Write a short product review for a {{ product_category }} item."\n )\n)
preview = data_designer.preview(config_builder=config)\npreview.display_sample_record()\n```
This release also incorporates the synthetic data tech my team originally built at Gretel (now part of NVIDIA), now generally available for anyone to use or extend.
Repo: https://github.com/NVIDIA-NeMo/DataDesigner"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia open sources the synthetic data framework used to build Nemotron datasets"}},"_tags":["story","author_alexwatson405","story_46136055","ask_hn"],"author":"alexwatson405","children":[46136161],"created_at":"2025-12-03T16:02:01Z","created_at_i":1764777721,"num_comments":1,"objectID":"46136055","points":8,"story_id":46136055,"story_text":"NVIDIA just open sourced NeMo Data Designer, the synthetic data framework used internally to build both pre-training and post-training datasets for Nemotron.
It lets you define an entire synthetic data pipeline directly in Python: structured outputs, statistical samplers, LLM-generated columns, dependency-aware field relationships, Python/SQL/remote validators, and optional LLM-as-judge scoring. Supports quick preview mode for fast iteration before scaling up.
Install:
```\npip install data-designer\n```
A minimal example:
```\nfrom data_designer.essentials import *
data_designer = DataDesigner()\nconfig = DataDesignerConfigBuilder()
config.add_column(\n SamplerColumnConfig(\n name="product_category",\n sampler_type=SamplerType.CATEGORY,\n params=CategorySamplerParams(\n values=["Electronics", "Clothing", "Home & Kitchen", "Books"]\n ),\n )\n)
config.add_column(\n LLMTextColumnConfig(\n name="review",\n model_alias="nvidia-text",\n prompt="Write a short product review for a {{ product_category }} item."\n )\n)
preview = data_designer.preview(config_builder=config)\npreview.display_sample_record()\n```
This release also incorporates the synthetic data tech my team originally built at Gretel (now part of NVIDIA), now generally available for anyone to use or extend.
Repo: https://github.com/NVIDIA-NeMo/DataDesigner","title":"Nvidia open sources the synthetic data framework used to build Nemotron datasets","updated_at":"2026-03-05T23:06:33Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bcatanzaro"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia Nemotron Nano v2"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://research.nvidia.com/labs/adlr/NVIDIA-Nemotron-Nano-2/"}},"_tags":["story","author_bcatanzaro","story_44943524"],"author":"bcatanzaro","children":[44946528],"created_at":"2025-08-18T18:04:14Z","created_at_i":1755540254,"num_comments":1,"objectID":"44943524","points":8,"story_id":44943524,"title":"Nvidia Nemotron Nano v2","updated_at":"2026-03-05T22:30:48Z","url":"https://research.nvidia.com/labs/adlr/NVIDIA-Nemotron-Nano-2/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ksec"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"Nvidia Debuts Nemotron 3 Family of Open Models"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nemotron"],"value":"https://nvidianews.nvidia.com/news/nvidia-debuts-nemotron-3-family-of-open-models"}},"_tags":["story","author_ksec","story_46275709"],"author":"ksec","created_at":"2025-12-15T15:27:38Z","created_at_i":1765812458,"num_comments":0,"objectID":"46275709","points":8,"story_id":46275709,"title":"Nvidia Debuts Nemotron 3 Family of Open Models","updated_at":"2026-03-05T23:12:31Z","url":"https://nvidianews.nvidia.com/news/nvidia-debuts-nemotron-3-family-of-open-models"}],"hitsPerPage":20,"nbHits":868,"nbPages":44,"page":0,"params":"query=Nemotron&tags=story&advancedSyntax=true&analyticsTags=backend","processingTimeMS":8,"processingTimingsMS":{"_request":{"queue":86,"roundTrip":15},"afterFetch":{"merge":{"mergeLoop":{"prepareNextHit":1,"total":1},"total":1},"total":1},"fetch":{"query":5,"total":6},"total":8},"query":"Nemotron","serverTimeMS":95}