# NVIDIA Chain-of-Thought Math & Code Reasoner (nvidia-nemotron-deep-reasoning-r1) Pricing & Spec Oracle (`ai-model-spec-nvidia-nemotron-deep-reasoning-r1`)
- **Status**: 200 OK | **Latency**: 1ms | **Cached**: true
- **Primary Authority**: 100% Platform AI Model Benchmark Registry
- **License & Trust**: 100% Platform Deterministic Compute / Open Reference (computed_by_platform)
- **Verified At (UTC)**: 2026-10-27T16:25:37.170Z
- **Data Freshness**: cached
## Structured Data
```json
{
  "model_id": "nvidia-nemotron/deep-reasoning-r1",
  "provider": "NVIDIA",
  "tier": "Chain-of-Thought Math & Code Reasoner",
  "context_window_tokens": 128000,
  "max_output_tokens": 32000,
  "pricing_usd_per_1m_tokens": {
    "input": 1.12,
    "output": 3.36,
    "cached_input": 0.28
  },
  "supported_capabilities": "Deep Multi-Step Planning & Verification",
  "supports_mcp_tools": true,
  "workload_simulation": {
    "calls_per_day": 1000,
    "daily_cost_usd": 3.92,
    "monthly_cost_usd": 117.6
  }
}
```
> **AI Agent Hint**: Agent Data Hub: ask anything via GET /api/v1/ask?q=..., whole tasks via /api/v1/bundles, up to 10 calls via POST /api/v1/batch. Add compact=1 or fields=a,b to save tokens. Docs: /llms.txt