Nemotron 3 Nano Omni
LLM Provider / Model ยท Local omni-modal workflows (video, audio, document), edge deployment on consumer GPUs, multimodal RAG, on-device assistants
At a glance
| Input price | Free (self-hosted) |
| Output price | Free (self-hosted) |
| Price tier | Free |
| Context window | 256K |
| Max output | 16K |
| Context tier | 128K-500K |
| Speed tier | Fast |
| Latency | local-bound |
| Knowledge cutoff | Jan 2026 |
| Modality | Multimodal (vision + audio) |
| Model ID | nemotron-3-nano-omni-30b-a3b |
| Provider | NVIDIA |
| HumanEval | 78% |
| MMLU | 78% |
| Benchmark | Omni-modal: high |
| Released | Apr 2026 |
| Hosting | Open-weights |
| Capabilities | Vision, Audio, Video, Tool use, Streaming, Structured output, MoE (30B / 3B active), 9x throughput on video/document workloads, Runs on 25GB RAM, Self-hostable |
What Nemotron 3 Nano Omni does
Vision, Audio, Video, Tool use, Streaming, Structured output, MoE (30B / 3B active), 9x throughput on video/document workloads, Runs on 25GB RAM, Self-hostable
Best for
Local omni-modal workflows (video, audio, document), edge deployment on consumer GPUs, multimodal RAG, on-device assistants
Works well with
Agent / Orchestration
Conflicts & caveats
- Trae's built-in model menu may not include Nemotron 3 Nano Omni โ Trae primarily ships with Claude/GPT/Gemini/DeepSeek.
- Google Antigravity is built around Gemini 3 Pro; Claude and recent GPT-5 models also work, but Nemotron 3 Nano Omni is not in the supported model list.
- Kiro's spec mode is tuned for Claude models โ Nemotron 3 Nano Omni may work but spec-driven flows are best with Claude.
- Cursor's built-in model menu may not include Nemotron 3 Nano Omni โ verify it's available in your Cursor model picker.
- OpenAI Codex (VS Code) is tied to OpenAI's GPT-5.1 Codex family โ Nemotron 3 Nano Omni is not selectable inside this integration.
- Gemini CLI is locked to Google Gemini models โ your LLM is Nemotron 3 Nano Omni.
- Privacy conflict: Self-hosted Nemotron 3 Nano Omni sends code to cloud @codebase (Cursor). Use local context (Continue indexing, ChromaDB, LanceDB, pgvector, Vespa self-hosted) for true privacy.
- Claude Code only works with Claude models โ your LLM is Nemotron 3 Nano Omni.
- Devin uses Cognition's proprietary model internally โ your selected Nemotron 3 Nano Omni will be ignored.
- OpenAI Codex (Cloud) runs GPT-5.1 Codex internally โ your selected Nemotron 3 Nano Omni will be ignored.
- Jules runs Gemini models internally โ your selected Nemotron 3 Nano Omni will be ignored.
- The Claude Code SDK is tuned for Claude models โ Nemotron 3 Nano Omni may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
- Lovable uses a fixed internal model โ your selected Nemotron 3 Nano Omni will be ignored.
- Replit Agent 3 uses its own model selection internally โ your selected Nemotron 3 Nano Omni will be ignored.
Build a full stack around Nemotron 3 Nano Omni โ Flowpicker shows compatibility warnings before you commit.
Open the stack planner โ