DeepSWE
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
Context window on the 2026-08-19 OpenRouter row: 256,000 tokens.
Max completion tokens on that row: 65,536.
Input / output on that row: $0 / $0 per 1M tokens.
Architecture fields: text+image+audio+video->text · Other.
Hugging Face id on that row: nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16.
NVIDIA Nemotron™ 3 Nano Omni is a 30B-A3B open multimodal model designed to function as a perception and context sub-agent in enterprise agent systems.
When you want the 2026-04-28 catalog SKU, not a later rename.
No independent board lists this exact SKU as of 2026-08-19.
Closest benchmarked sibling: Nemotron 3 Ultra, 27.39% Vals Index on the vals.ai hero index board. That is a score for the sibling, not for this SKU.
Boards checked: DeepSWE official (2026-08-13), Artificial Analysis Intelligence Index, vals.ai hero index. Every card that does carry a row is on the coding leaderboard.
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
No Intelligence Index integer fetched for this identity. Chip omitted.
We did not open a vals.ai card for this identity. Chip omitted.
Plus is $25/mo with $25 weekly hosted usage. The Mac app stays free with your own keys. Get Plus is not Download for Mac.
Official weights id on the 2026-08-19 OpenRouter row: nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16. Community GGUF is a quant, not a second Continuum card.
This is the model repo, not Download for Mac. Downloads and likes from the Hugging Face API on 2026-08-19.
usedStorage and the sum of .safetensors sizes can differ (LFS pointers, non-shard files, duplicate copies). Cited separately. Source: recursive tree API, 2026-08-19.
| Path | Bytes |
|---|---|
model-00001-of-00017.safetensors | 3,996,912,760 |
model-00002-of-00017.safetensors | 3,999,538,784 |
model-00003-of-00017.safetensors | 3,994,808,632 |
model-00004-of-00017.safetensors | 3,999,538,936 |
model-00005-of-00017.safetensors | 3,994,809,040 |
model-00006-of-00017.safetensors | 3,999,539,160 |
model-00007-of-00017.safetensors | 3,994,809,064 |
model-00008-of-00017.safetensors | 3,999,539,152 |
model-00009-of-00017.safetensors | 3,994,809,088 |
model-00010-of-00017.safetensors | 3,999,539,128 |
model-00011-of-00017.safetensors | 3,982,766,872 |
model-00012-of-00017.safetensors | 3,991,625,352 |
model-00013-of-00017.safetensors | 3,970,314,024 |
model-00014-of-00017.safetensors | 3,994,100,200 |
model-00015-of-00017.safetensors | 3,999,539,368 |
model-00016-of-00017.safetensors | 3,997,900,128 |
model-00017-of-00017.safetensors | 2,122,218,848 |
tokenizer.json | 17,077,367 |
media/demo.mp4 | 7,333,182 |
model.safetensors.index.json | 819,092 |
media/2414-165385-0000.wav | 638,798 |
media/tech.png | 222,054 |
tokenizer_config.json | 188,045 |
media/table.png | 131,014 |
modeling_nemotron_h.py | 64,337 |
README.md | 48,015 |
modeling.py | 28,713 |
processing.py | 26,284 |
media/example1a.jpeg | 14,890 |
chat_template.jinja | 14,277 |
configuration_nemotron_h.py | 13,971 |
media/example1b.jpeg | 12,075 |
image_processing.py | 10,883 |
config.json | 9,842 |
configuration_radio.py | 8,982 |
audio_model.py | 6,888 |
video_processing.py | 6,530 |
video_io.py | 6,523 |
configuration.py | 4,872 |
explainability.md | 4,792 |
processing_utils.py | 3,018 |
Showing the first 17 of 17 safetensor shards, plus 24 other files. The snapshot was capped at 17 shards when the tree API was read on 2026-08-19, so the remaining 0 are on the Files tab, not on this page. Totals above are the full tree API sum, not this slice. Byte sizes are Hub tree API size fields, never invented.
OpenRouter slug nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free on the 2026-08-19 catalog.
First party: https://build.nvidia.com/.
Not on the Continuum host list fetched 2026-08-19.
We list first-party and OpenRouter, plus Continuum only when the live public allowlist named the id. This is not a 15-host routing table.
OpenRouter 2026-08-19: $0 in / $0 out per 1M tokens. Context 256,000 in / 65,536 out.
No separate internal-reasoning price on that row.
| Claim | Source | As of |
|---|---|---|
| OpenRouter id nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free; context 256000; created 1777393095; $0 / $0 per 1M | OpenRouter /api/v1/models | 2026-08-19 |
| OpenRouter id nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free, context 256,000 | OpenRouter /api/v1/models | 2026-08-19 |
| HF downloads 405,584 | Hugging Face API nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 2026-08-19 |
| HF safetensors.total: 33,015,632,214 stored tensors; HF API usedStorage 66,057,737,916 bytes; created 2026-04-20T04:40:42.000Z; HF tensors F32 + BF16 | Hugging Face API nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 2026-08-19 |
| 17 safetensor shards; shard bytes 66,032,308,536; tree files 50 | Hugging Face tree API nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 2026-08-19 |
No other card in this cluster shares a published DeepSWE official row we can put next to this one. We will not compare on vendor-blog numbers.
Continue through NVIDIA's model family with Nemotron 3 Nano 30B A3B.
Lab posts pick a harness, an effort, a split, and sometimes a private eval set. DeepSWE publishes the official mini-swe-agent row with cost, tokens, and steps. Scale SWE-bench Pro only counts when the public shared-harness board has a row we can fetch. A higher lab number is usually a different test, not a better one.
No Continuum hosted id is verified for this model, so this page does not invent a curl target. Use the first-party API or the OpenRouter slug nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free.