Excited to support @NVIDIA Nemotron 3 Nano Omni, now available on Fireworks. It's the first open mo...
TL;DR · AI 摘要
Fireworks AI 宣布上线 NVIDIA Nemotron 3 Nano Omni——首个开源多模态模型，支持图文音视单次推理，256K上下文，吞吐量为Qwen3 30B的9倍。
核心要点
- Nemotron 3 Nano Omni是首个开源的统一处理文本、图像、音频、视频的单循环推理模型
- 专为大规模多模态子智能体（sub-agents）设计，强调工程可部署性
- 在Fireworks平台提供按需部署，但未披露架构细节、评测方法或开源协议
结构提纲
按章节快速跳转。
- §发布通告
Fireworks AI宣布支持NVIDIA Nemotron 3 Nano Omni模型上线。
- ·核心宣称
强调其‘首个开源’‘四模态单循环’‘256K上下文’‘9×吞吐优势’四大卖点。
- ·部署信息
模型已集成至Fireworks平台，支持on-demand推理服务。
- ›技术留白
未说明模型参数量、训练数据、开源许可证、基准评测或实际延迟指标。
思维导图
用一张图看清主题之间的关系。
查看大纲文本（无障碍 / 无 JS 友好）
- Nemotron 3 Nano Omni
- 多模态能力
- 文本/图像/音频/视频单循环
- 性能指标
- 9× Qwen3 30B吞吐
- 256K上下文
- 部署形态
- Fireworks on-demand服务
金句 / Highlights
值得收藏与分享的关键句。
It's the first open model that handles vision, audio, video, and text in a single inference loop.
Built for multimodal sub-agents at scale, with 9× higher throughput than Qwen3 30B.
256K context.
Now available on Fireworks for on-demand deployments.
It's the first open model that handles vision, audio, video, and text in a single inference loop. Built for multimodal sub-agents at scale, with 9× higher throughput than Qwen3 30B. 256K context.
Now https://t.co/EqWcuATF0W" / X
Fireworks AI on X: "Excited to support @NVIDIA Nemotron 3 Nano Omni, now available on Fireworks. It's the first open model that handles vision, audio, video, and text in a single inference loop. Built for multimodal sub-agents at scale, with 9× higher throughput than Qwen3 30B. 256K context. Now https://t.co/EqWcuATF0W" / X
Don’t miss what’s happening
Excited to support
Nemotron 3 Nano Omni, now available on Fireworks. It's the first open model that handles vision, audio, video, and text in a single inference loop. Built for multimodal sub-agents at scale, with 9× higher throughput than Qwen3 30B. 256K context. Now available on Fireworks for on-demand deployments. →https://fireworks.ai/models/firewor ks/nvidia-nemotron-3-nano-omni-30b-a3b…
Last edited Opens edit history 4:09 PM · Apr 28, 2026
·
2
3
57
5