NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16
--- libraryname: transformers license: other licensename: openmdw-1.1 licenselink: https://openmdw.ai/license/1-1/ pipelinetag: text-generation language: - en - es - fr - de - it - ja tags: - nvidia - pytorch - nemotron-3.5 datasets: - nvidia/nemotron-post-training-v3 - nvidia/nemotron-pre-training-datasets trackdownloads: true ---
Params
31.6 B
Context
262,144
Downloads 30d
1.1 M
Likes
271
Download history
daily snapshots · 76 days
▲ 471 K in the last 30 days (76.9%)
1.2 M615 K
Oct 12Oct 22Nov 1Nov 10
1.2 M364 K
Aug 27Sep 21Oct 16Nov 10
Can you run it?
Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.
| File | Quant | Size | Est. VRAM | Verdict on RTX 4090 · 24 GB |
|---|---|---|---|---|
| model.safetensors | bf16 | 65.8 GB | 77.6 GB | ❌ Won’t fit |
| model.safetensors (bf16, full) | bf16 + 256K ctx | 65.8 GB | 224.5 GB | ❌ Won’t fit |
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.
Run it
copy-paste, exact tags checked against the Hub$ curl -s https://aimodelscomparison.com/api/v1/models/nvidia-nemotron-3-5-lightning-30b-a3b-bf16
{
"hf_id": "nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16",
"params_b": 31.58,
"context_length": 262144,
"license": { "id": "other", "commercial": "unknown" },
"downloads_30d": 1082506,
"vram_estimates": [
{ "quant": "bf16", "gb": 77.6 }
],
"updated_at": "2026-08-27T01:00:14Z"
}
Use via API
via OpenRouter · USD per 1M tokens · synced 2026-11-10$0.07
input / 1M
$0.20
output / 1M
$0.04
cached input / 1M
$0.27
1M in + 1M out
:free variant
Free in · Free out
- Context window
- 262,144 tokens
- Max output
- 235,929 tokens
- Providers
- 5
- Cheapest provider
- Darkbloom · $0.025 / $0.12
- Best throughput
- —
- AA Intelligence / Coding / Agentic
- 5.5 / 26.8 / 3.5
✓ Tool calling
✓ Reasoning
✓ Structured outputs
✗ Image input
✗ File input
✗ Audio input
✓ Prompt caching
Specifications
- Architecture
- NemotronHForCausalLM
- Parameters
- 31.6 B
- Tensor type
- BF16
- Context length
- 262,144
- Vocabulary
- 131,072
- Layers / heads
- 52 / 32
- Licence
- other
- First seen on the Hub
- 2026-08-01
- Training datasets
- nvidia/nemotron-post-training-v3, nvidia/nemotron-pre-training-datasets
- Added to our catalog
- 2026-08-27
Family
Base model and the most-downloaded derivatives in the catalog.
Compare with any text-generation model