DeepSWE
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
nvidia/nemotron-3-super-120b-a12b:freeContext window on the 2026-08-19 OpenRouter row: 262,144 tokens.
Max completion tokens on that row: 262,144.
Input / output on that row: $0 / $0 per 1M tokens.
Architecture fields: text->text · Other.
Hugging Face id on that row: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8.
NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications.
When you want the 2026-03-11 catalog SKU, not a later rename.
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
We did not open a vals.ai card for this identity. Chip omitted.
Ranked view: Best coding models: the independent leaderboard puts this row and every other card that carries an independent coding score on one board, with the confidence intervals left visible.
Copied from AA or vals HTML we opened. 0.0% placeholder bars are omitted. Do not average these into the hero tiles, and do not invent CI, $/task, or steps for Artificial Analysis.
| Bench | Printed | Note | Source | URL | As of |
|---|---|---|---|---|---|
| GDPval-AA v2 | 9.9% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| τ³-Banking | 10.3% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| Terminal-Bench v2.1 | 38.6% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| SciCode | 36% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| Humanity's Last Exam | 20.8% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| GPQA Diamond | 80% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| CritPt | 3.1% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| AA-Omniscience Accuracy | 24.3% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
| AA-LCR | 60.3% | Printed on the AA Intelligence Evaluations grid for Nemotron 3 Super 120B A12B (Reasoning). Not a DeepSWE chip. | Artificial Analysis Nemotron 3 Super | source | 2026-08-19 |
nvidia/nemotron-3-super-120b-a12b:free. Use a different card on this hub when you need another lab SKU.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/NVIDIA-Nemotron-3-Super-120B-A12B-FP8. 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-00026.safetensors | 5,000,243,808 |
model-00002-of-00026.safetensors | 4,993,713,376 |
model-00003-of-00026.safetensors | 5,000,119,040 |
model-00004-of-00026.safetensors | 4,999,076,840 |
model-00005-of-00026.safetensors | 5,000,270,768 |
model-00006-of-00026.safetensors | 4,999,077,288 |
model-00007-of-00026.safetensors | 4,998,813,984 |
model-00008-of-00026.safetensors | 5,000,533,616 |
model-00009-of-00026.safetensors | 4,999,077,224 |
model-00010-of-00026.safetensors | 4,998,813,992 |
model-00011-of-00026.safetensors | 5,000,533,608 |
model-00012-of-00026.safetensors | 4,999,077,168 |
model-00013-of-00026.safetensors | 4,998,813,984 |
model-00014-of-00026.safetensors | 5,000,533,616 |
model-00015-of-00026.safetensors | 4,999,077,104 |
model-00016-of-00026.safetensors | 4,998,814,008 |
model-00017-of-00026.safetensors | 5,000,533,600 |
model-00018-of-00026.safetensors | 4,999,077,048 |
model-00019-of-00026.safetensors | 4,998,814,040 |
model-00020-of-00026.safetensors | 5,000,533,560 |
model-00021-of-00026.safetensors | 4,999,076,984 |
model-00022-of-00026.safetensors | 4,998,814,112 |
model-00023-of-00026.safetensors | 5,000,533,512 |
model-00024-of-00026.safetensors | 4,999,076,944 |
model-00025-of-00026.safetensors | 4,998,841,656 |
model-00026-of-00026.safetensors | 3,368,110,800 |
tokenizer.json | 17,077,484 |
model.safetensors.index.json | 12,390,752 |
tokenizer_config.json | 177,209 |
modeling_nemotron_h.py | 82,338 |
accuracy_chart.png | 79,162 |
README.md | 79,162 |
configuration_nemotron_h.py | 19,822 |
chat_template.jinja | 10,771 |
config.json | 8,440 |
hf_quant_config.json | 6,888 |
explainability.md | 3,155 |
privacy.md | 2,688 |
bias.md | 2,626 |
safety.md | 2,120 |
super_v3_reasoning_parser.py | 1,878 |
.gitattributes | 1,635 |
special_tokens_map.json | 563 |
generation_config.json | 210 |
__init__.py | 0 |
All 45 files the tree API returned are listed: every one of the 26 safetensor shards plus 19 other files. Files tab: https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8/tree/main.
OpenRouter slug nvidia/nemotron-3-super-120b-a12b:free on the 2026-08-19 catalog.
First party: https://build.nvidia.com/.
Continuum hosts nvidia/nemotron-3-super-120b-a12b:free on the live public allowlist fetched 2026-08-19.
nvidia/nemotron-3-super-120b-a12b:free
click to select
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 262,144 in / 262,144 out.
No separate internal-reasoning price on that row.
| Claim | Source | As of |
|---|---|---|
| OpenRouter id nvidia/nemotron-3-super-120b-a12b:free; context 262144; created 1773245239; $0 / $0 per 1M | OpenRouter /api/v1/models | 2026-08-19 |
| AA Intelligence Index 26; 9 printed Index benches | Artificial Analysis model page | 2026-08-19 |
| OpenRouter id nvidia/nemotron-3-super-120b-a12b:free, context 262,144 | OpenRouter /api/v1/models | 2026-08-19 |
| HF downloads 175,033 | Hugging Face API nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8 | 2026-08-19 |
| HF safetensors.total: 123,611,012,096 stored tensors; HF API usedStorage 128,379,469,916 bytes; created 2026-03-10T18:32:42.000Z; HF tensors F32 + BF16 + F8_E4M3 | Hugging Face API nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8 | 2026-08-19 |
| 26 safetensor shards; shard bytes 128,350,001,680; tree files 45 | Hugging Face tree API nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8 | 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 Omni (free).
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.
Verified against the live public allowlist on 2026-08-19. Base URL is https://continuumcode.ai/v1. Keys are cont_sk_ from Settings, Account, Inference API. Personal keys need Plus or above.
curl https://continuumcode.ai/v1/chat/completions \
-H "Authorization: Bearer $CONTINUUM_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"nvidia/nemotron-3-super-120b-a12b:free","messages":[{"role":"user","content":"Review this diff."}]}'