{
  "schemaVersion": "1.0.0",
  "generatedAt": "2026-08-08T19:13:47.292Z",
  "mode": "live-tavily-keyless-search-local-synthesis",
  "disclosure": "Tavily search was live. The final line was deterministic local synthesis; no Token Factory request was made.",
  "query": "Nebius Token Factory function calling MCP agent reliability",
  "sources": [
    {
      "title": "Function calling & Tools - Nebius Token Factory documentation",
      "url": "https://docs.tokenfactory.nebius.com/ai-models-inference/function-calling",
      "content": "Function calling allows you to enhance a model’s capabilities by integrating it with external code, services, or APIs. You can define a set of functions—also referred to as tools—that the model can call. Based on the context provided in the conversation, the model can intelligently decide which functions to invoke and suggest appropriate arguments. This capability enables you to build more powerful agents that can: [...] Run custom code\n Retrieve real-time data\n Interact with third-party services\n Connect to MCP servers\n And perform other dynamic actions\n\nImportant: The model does not execute the functions itself. Instead, it outputs tool invocation instructions in an OpenAI-compatible function calling format. You are responsible for parsing these outputs and executing the corresponding functions on your backend or client side.\n\n1\n\nDefine Available Tools [...] If a function call is proposed, your system is responsible for executing the function with the provided arguments. You then pass the function’s output back to the model as part of the next prompt. This allows the conversation to continue seamlessly, with the model incorporating the function result into its response or trigger",
      "score": 0.6922474
    },
    {
      "title": "Nebius Token Factory",
      "url": "https://nebius.com/services/token-factory",
      "content": "## AI agent essentials\n\nBuild and deploy intelligent agents faster with native function calling, structured JSON outputs and built-in safety guardrails for reliable real-world interaction.\n\n## Custom and fine-tuned models\n\nAdapt models to your data using fine-tuning workflows. Deploy your own checkpoints directly on Token Factory endpoints with guaranteed performance and transparent per-token pricing.\n\n## RAG development tools [...] Sub-second responses and stable latency, even at peak load. Top tier performance on models like DeepSeek V3 0324, independently verified by Artificial Analysis.\n\n### Scale without limits\n\nHandle 100M+ tokens per minute with consistent throughput and 99.9% uptime SLAs. Autoscaling and speculative decoding ensure reliability from prototype to global deployment.\n\n### Comprehensive model coverage\n\nAccess 60+ premium models spanning LLMs, vision, reasoning and embeddings, expanding monthly. [...] ## Questions and answers about Nebius Token Factory\n\nCan I use your service for large production workloads?\n\nYes. Nebius Token Factory is built for large-scale, production-grade AI workloads.\n\nDedicated endpoints deliver sub-second inference, 99.9% uptime, and autos",
      "score": 0.608897
    },
    {
      "title": "Agent 101: Launching production-grade agents at scale",
      "url": "https://nebius.com/blog/posts/launch-production-agents-at-scale",
      "content": "Nebius MCP example: Learn how to spin a Model Context Protocol (MCP) for your agents by using Nebius.\n\n# Building production-ready agentsBuilding production-ready agents\n\nPrototyping an agent is easy, but making it production-ready is where the real engineering begins.\n\nIn this article, we’ve walked through what it takes to evolve from demo agents to production-grade systems that are reliable under load, observable in real time and modular enough to adapt as needs change. [...] To provide agents with access to the latest information from the web, you can use search tools like LinkUp. LinkUp is a web search API specifically designed to connect AI applications to the internet, providing real-time, factual information.\n\nEnable real-time web search in your AI agents by integrating LinkUp through function calling. This setup works seamlessly with all text generation models on Nebius, allowing agents to retrieve up-to-date, factual information when needed. [...] # Agent 101: Launching production-grade agents at scale\n\nAgent 101: Launching production-grade agents at scale\n\nNebius AI Studio is now Nebius Token Factory: same platform, new name, more power for running AI at scale.\n\n# IntroIn",
      "score": 0.47139415
    }
  ],
  "trace": [
    {
      "node": "search_web",
      "provider": "Tavily",
      "resultCount": 5,
      "requestId": "491bdba8-9050-4dc7-9a42-61403b8608f0",
      "usage": {
        "credits": 1
      }
    },
    {
      "node": "select_evidence",
      "selectedCount": 3
    },
    {
      "node": "answer_with_sources",
      "provider": "Token Factory compatible answerer",
      "usage": null,
      "estimatedCostUsd": 0
    }
  ],
  "answer": "Search-only smoke completed with 3 selected sources. Configure NEBIUS_API_KEY and a current model to use the Token Factory answerer."
}