{
  "key": "qwen3-8-2-4t-a95b-nvfp4",
  "name": "Qwen3.8 2.4T A95B (NVFP4)",
  "description": "Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows",
  "lab": "alibaba",
  "family": "qwen",
  "release_date": "2026-08-12",
  "knowledge": null,
  "open_weights": true,
  "has_free_offering": false,
  "published": true,
  "modalities": {
    "input": [
      "text"
    ],
    "output": [
      "text"
    ]
  },
  "capabilities": {
    "reasoning": true,
    "tool_call": true,
    "structured_output": true,
    "attachment": false,
    "reasoning_options": [
      {
        "type": "effort",
        "values": [
          "low",
          "medium",
          "xhigh"
        ]
      }
    ]
  },
  "limit": {
    "context": 262144,
    "output": 32768
  },
  "offerings": [
    {
      "provider": "runinfra",
      "provider_model_id": "Inferact/Qwen3.8-2.4T-A95B-NVFP4",
      "name": "Qwen3.8 2.4T A95B (NVFP4)",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 2,
        "output": 6,
        "cache_read": 0.2,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 262144,
        "output": 32768,
        "input": null
      },
      "status": null,
      "last_updated": "2026-08-12"
    }
  ]
}