比较服务商

每个点是一家服务商。从左到右是每百万 token 的输入价,从下到上是输出价。

基础价格

9
最新价格检查于 Nov 18, 2026。

Nemotron 3.5 Super VL · 已公布价格

Trial and free endpoint pricing status
Billing categoryInput / 1MOutput / 1MCache read / 1M
NVIDIA hosted trialNo paid tariff publishedNo paid tariff publishedNot reported
OpenRouter NVIDIA :free$0$0Not reported
[2][3]

Nemotron 3.5 Super VL · 权重许可证

权重许可证
模型权重协议
开放权重OpenMDW 1.1
[1]

Nemotron 3.5 Super VL · 权重与量化

Local deployment and enterprise agents
ItemConfiguration / evidence
CheckpointNemotron 3.5 Super VL 120B-A12B NVFP4; OpenMDW 1.1
QuantizationW4A4 on Blackwell; W4A16 on non-Blackwell, per serving configuration
[2]

Nemotron 3.5 Super VL · 自部署硬件

Local deployment and enterprise agents
ItemConfiguration / evidence
Supported hardwareBlackwell GB200 / DGX Spark; Hopper H100 / H200
[2]

Nemotron 3.5 Super VL · 部署方式

Local deployment and enterprise agents
ItemConfiguration / evidence
Serving enginesvLLM / SGLang; NVIDIA GPU-accelerated systems
[2]

Nemotron 3.5 Super VL · 上下文与渠道限制

Local deployment and enterprise agents
ItemConfiguration / evidence
Local contextUp to 1M; recipes may reduce context for GPU memory and concurrency
[2]

Nemotron 3.5 Super VL · 已知风险与限制

Local deployment and enterprise agents
ItemConfiguration / evidence
Benchmark scopeNVIDIA evaluations use NeMo Gym recipes; not independent LMSpeed measurements
Model identityThis is Super VL 120B-A12B; no BOLT variant or edge latency result verified
[2]

Nemotron 3.5 Super VL · 使用场景

Local deployment and enterprise agents
ItemConfiguration / evidence
Enterprise use casesDocument and video understanding; coding, reasoning and tool-using agents
[2]

上下文与限制

  • It does not publish a paid per-token tariff.

    [2][3]
  • NVIDIA Nemotron 3.5 Super VL has 120 billion total parameters and 12 billion active parameters.

    [2]
  • The model card currently describes the NVFP4 quantized release.

    [2]
  • Use of the model is governed by the OpenMDW License Agreement version 1.1.

    [2][3]
  • Access to the hosted trial is governed by the NVIDIA API Trial Terms of Service.

    [2][3]

关于 Nemotron 3.5 Super VL

来源与注脚

  1. [1]
    来源核验日期:
  2. [2]
  3. [3]
    OpenRouter NVIDIA free endpoint核验日期:

输入、输出和缓存读取的单位是每百万 token 的美元价格。速度是每秒输出 token 数。可用率是来源报告的在线比例。