{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"nvidia-nim","slug":"nvidia-nim","name":"NVIDIA NIM","type":"platform","url":"https://build.nvidia.com","page_url":"https://unfragile.ai/nvidia-nim","categories":["deployment-infra"],"tags":[],"pricing":{"model":"freemium","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"nvidia-nim__cap_0","uri":"capability://text.generation.language.openai.compatible.inference.api.with.multi.model.routing","name":"openai-compatible inference api with multi-model routing","description":"Exposes NVIDIA NIM-optimized models through OpenAI API-compatible endpoints (e.g., /v1/chat/completions, /v1/completions), enabling drop-in replacement of OpenAI clients without code changes. Routes requests to containerized TensorRT-LLM inference engines running on NVIDIA GPUs, with automatic model selection from a curated catalog including DeepSeek-v4-pro, Nemotron-3-nano-omni, GLM-5.1, and Gemma-4-31b-it. Supports text generation and reasoning tasks through standardized request/response payloads.","intents":["Replace OpenAI API calls with NVIDIA-optimized inference without refactoring client code","Route inference requests to specific NVIDIA-optimized model versions","Evaluate multiple reasoning and language models through a single API interface","Integrate NVIDIA GPU-optimized inference into existing LLM applications"],"best_for":["Teams migrating from OpenAI to on-premises or edge inference","Developers building multi-model applications requiring API consistency","Enterprises requiring inference on NVIDIA hardware for compliance or performance"],"limitations":["API compatibility is claimed but not verified in source material — exact endpoint paths and payload structures unknown","Model availability limited to NVIDIA-curated catalog; custom model deployment requirements unknown","Streaming, batch, and async response modes not documented in available material","Rate limiting, quota, and token limits per model not specified"],"requires":["NVIDIA API key (authentication mechanism unconfirmed)","OpenAI-compatible client library (Python, Node.js, etc.)","Network access to NVIDIA NIM deployment (cloud, on-prem, or edge)"],"input_types":["text (chat messages, prompts)","structured JSON (OpenAI chat completion format)"],"output_types":["text (model completions)","structured JSON (OpenAI response format with tokens, finish_reason)"],"categories":["text-generation-language","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_1","uri":"capability://automation.workflow.tensorrt.llm.optimized.inference.container.deployment","name":"tensorrt-llm optimized inference container deployment","description":"Packages pre-optimized inference engines using NVIDIA's TensorRT-LLM framework into containerized microservices that can be deployed across cloud, on-premises, and edge environments. Each container includes model weights, quantization profiles, and kernel optimizations targeting specific NVIDIA GPU architectures (Blackwell B300/B200, Hopper H200, RTX Pro 6000). Deployment abstracts hardware-specific optimization details, exposing a unified inference interface regardless of target infrastructure.","intents":["Deploy inference workloads to on-premises NVIDIA GPU clusters without manual TensorRT-LLM tuning","Run the same inference container across multiple deployment environments (cloud, edge, data center)","Achieve maximum inference throughput on NVIDIA hardware without custom optimization","Reduce inference latency through GPU-specific kernel compilation and quantization"],"best_for":["Enterprise teams deploying inference on owned NVIDIA GPU infrastructure","Organizations with strict data residency or compliance requirements preventing cloud inference","Edge AI deployments requiring optimized inference on RTX or Jetson hardware"],"limitations":["Requires NVIDIA GPU hardware; no CPU-only inference option documented","Supported GPU models limited to Blackwell (B300, B200), Hopper (H200), and RTX Pro 6000 — compatibility with older architectures unknown","Container orchestration requirements (Kubernetes, Docker) not documented","Model quantization profiles and optimization levels not configurable in source material","Inference performance benchmarks and SLA guarantees not provided"],"requires":["NVIDIA GPU (B300, B200, H200, or RTX Pro 6000 minimum)","NVIDIA CUDA runtime and cuDNN libraries","Container runtime (Docker or Kubernetes)","Network connectivity for model serving (port configuration unknown)"],"input_types":["container image (pre-built NVIDIA NIM container)","model weights (included in container)","inference requests (text prompts via API)"],"output_types":["inference results (text completions)","performance metrics (latency, throughput — format unknown)"],"categories":["automation-workflow","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_10","uri":"capability://automation.workflow.multi.gpu.and.distributed.inference.scaling","name":"multi-gpu and distributed inference scaling","description":"Supports distributed inference across multiple NVIDIA GPUs within a single deployment or across GPU clusters, enabling horizontal scaling for high-throughput inference workloads. Handles request batching, load balancing, and GPU memory management across multiple devices. Enables inference on models larger than single-GPU memory by distributing model weights and computation across GPUs.","intents":["Scale inference throughput across multiple GPUs for high-volume workloads","Deploy large models exceeding single-GPU memory by distributing across multiple devices","Implement load balancing and request batching across GPU cluster","Achieve high availability through multi-GPU redundancy"],"best_for":["Organizations deploying inference at scale with high throughput requirements","Teams running large models requiring multi-GPU distribution","Enterprises building high-availability inference infrastructure"],"limitations":["Multi-GPU scaling configuration and setup not documented","Load balancing strategy and request batching behavior not specified","Distributed inference latency overhead not quantified","GPU communication protocol (NVLink, PCIe, network) not documented","Maximum cluster size and scaling limits unknown","Failover and redundancy behavior not specified"],"requires":["Multiple NVIDIA GPUs (quantity and architecture requirements unknown)","GPU interconnect (NVLink for optimal performance, or network for distributed clusters)","Container orchestration supporting multi-GPU allocation (Kubernetes with NVIDIA device plugin, or Docker Compose)","Sufficient GPU memory aggregate for target model"],"input_types":["inference request (text prompt)","scaling configuration (number of GPUs, batching strategy — format unknown)"],"output_types":["inference result (text completion)","performance metrics (throughput, latency — format unknown)"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_11","uri":"capability://tool.use.integration.freemium.api.access.with.usage.based.pricing","name":"freemium api access with usage-based pricing","description":"Offers freemium access to NIM inference APIs, enabling developers to evaluate models and build prototypes without upfront cost. Free tier includes limited inference quota (exact limits unknown). Paid tiers scale with usage, with pricing based on inference volume or tokens consumed (pricing structure not documented). Enables cost-effective evaluation and gradual scaling from prototype to production.","intents":["Evaluate NVIDIA NIM models and API without upfront cost","Build and test prototypes using production inference infrastructure","Scale inference costs gradually as application usage grows","Compare NVIDIA NIM pricing against OpenAI or other inference APIs"],"best_for":["Developers prototyping AI applications","Teams evaluating NVIDIA NIM before committing to production deployment","Startups and small teams with limited inference budgets"],"limitations":["Free tier quota limits not documented — unclear how much inference is included","Paid tier pricing structure not provided — per-token, per-request, or subscription model unknown","Pricing comparison to OpenAI, Anthropic, or other APIs not available","Volume discounts or enterprise pricing not documented","Free tier rate limits and throttling behavior unknown","Billing and payment methods not specified"],"requires":["NVIDIA account (signup process unknown)","API key for authentication","Payment method for paid tier (if exceeding free quota)"],"input_types":["inference request (text prompt)","API key (authentication)"],"output_types":["inference result (text completion)","usage metrics and billing information (format unknown)"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_2","uri":"capability://automation.workflow.multi.environment.deployment.abstraction.cloud.on.premises.edge","name":"multi-environment deployment abstraction (cloud, on-premises, edge)","description":"Abstracts deployment infrastructure differences through a unified container interface, allowing the same NIM microservice to run on NVIDIA cloud platforms, on-premises data centers, or edge devices without code or configuration changes. Handles environment-specific resource allocation, networking, and GPU binding transparently. Supports DGX Station integration for on-premises enterprise deployments and edge inference on RTX hardware.","intents":["Deploy inference to cloud, on-prem, and edge without maintaining separate inference codebases","Migrate inference workloads between environments (e.g., cloud to on-prem) without re-optimization","Run inference on edge devices (RTX hardware) with the same API as cloud deployments","Integrate inference into existing DGX Station or enterprise GPU infrastructure"],"best_for":["Enterprises with hybrid cloud/on-prem infrastructure requiring unified inference deployment","Organizations deploying inference across multiple geographic regions or edge locations","Teams building AI applications requiring flexibility to shift inference location based on cost or latency"],"limitations":["Cloud provider options and regions not documented — unclear which cloud platforms are supported","On-premises deployment requirements (network, storage, compute) not specified","Edge device compatibility limited to RTX Pro 6000 — broader edge hardware support unknown","Cross-environment failover, load balancing, and traffic routing not documented","Pricing differences between deployment environments not provided"],"requires":["Target deployment environment (cloud account, on-prem GPU cluster, or edge device)","NVIDIA GPU hardware matching supported architectures","Container orchestration platform (Kubernetes for cloud/on-prem, Docker for edge)","Network connectivity between deployment environment and client applications"],"input_types":["deployment configuration (environment type, GPU allocation — format unknown)","container image (NVIDIA NIM container)","inference requests (API calls)"],"output_types":["deployed inference service (running container)","inference results (text completions)","deployment status and metrics (format unknown)"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_3","uri":"capability://memory.knowledge.curated.model.catalog.with.pre.optimized.weights","name":"curated model catalog with pre-optimized weights","description":"Maintains a curated selection of AI models (DeepSeek-v4-pro, Nemotron-3-nano-omni-30b-a3b-reasoning, GLM-5.1, Gemma-4-31b-it, and others) with pre-compiled TensorRT-LLM weights, quantization profiles, and GPU-specific optimizations. Each model is tested and validated on NVIDIA hardware, with documented capabilities (reasoning, text generation, OCR). Developers select models by name through the API without managing weights, quantization, or compilation.","intents":["Access production-ready models without downloading, compiling, or optimizing weights","Evaluate multiple models (reasoning, language, multimodal) through a single API","Deploy models with confidence that NVIDIA has validated performance on target GPU hardware","Use specialized models (e.g., Nemotron for reasoning) without custom integration"],"best_for":["Teams wanting to avoid model optimization and compilation complexity","Developers building applications requiring specific model capabilities (reasoning, OCR)","Organizations preferring NVIDIA-validated models over self-managed model deployments"],"limitations":["Model selection limited to NVIDIA's curated catalog — custom or fine-tuned models not supported (or requirements unknown)","Model availability may vary by deployment environment (cloud vs on-prem vs edge) — not documented","Model context windows, token limits, and training data not documented in source material","No information on model update frequency or version management","Exact number of available models unknown (homepage shows 4 explicitly, '+3' indicates more but unspecified)"],"requires":["NVIDIA NIM API access","Model name/identifier (e.g., 'deepseek-v4-pro')","Sufficient GPU memory for target model (model size requirements unknown)"],"input_types":["model identifier (string)","inference request (text prompt)"],"output_types":["model inference result (text completion)","model metadata (capabilities, parameters — format unknown)"],"categories":["memory-knowledge","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_4","uri":"capability://planning.reasoning.reasoning.specialized.model.inference.nemotron.3.nano.omni","name":"reasoning-specialized model inference (nemotron-3-nano-omni)","description":"Exposes NVIDIA's Nemotron-3-nano-omni-30b-a3b-reasoning model, a 30-billion-parameter model specifically trained for complex reasoning tasks, through the standard NIM API. The model is pre-optimized for TensorRT-LLM inference and supports chain-of-thought reasoning patterns. Enables applications requiring structured problem-solving, multi-step reasoning, or complex decision-making without requiring larger or more expensive reasoning models.","intents":["Perform complex reasoning tasks (math, logic, planning) on a 30B parameter model","Implement chain-of-thought reasoning patterns in applications without using larger models","Evaluate reasoning capabilities of a specialized model optimized for inference efficiency","Build agentic AI systems requiring structured reasoning on edge or on-premises hardware"],"best_for":["Teams building reasoning-heavy applications (math solvers, logic engines, planning systems)","Developers requiring reasoning capabilities on edge or on-premises hardware","Organizations evaluating reasoning model performance vs cost tradeoffs"],"limitations":["Model size (30B parameters) may exceed available GPU memory on smaller hardware (RTX 4090 or smaller — exact requirements unknown)","Reasoning performance benchmarks and accuracy metrics not provided","Training data, instruction tuning details, and reasoning capability boundaries not documented","Comparison to other reasoning models (e.g., o1, DeepSeek-R1) not provided","Context window and maximum reasoning depth not specified"],"requires":["NVIDIA GPU with sufficient VRAM (30B model size suggests 24GB+ — exact requirement unknown)","NVIDIA NIM API access","Model identifier: 'nemotron-3-nano-omni-30b-a3b-reasoning'"],"input_types":["text prompt (reasoning task, math problem, logic puzzle)","structured JSON (OpenAI chat completion format)"],"output_types":["text (reasoning response, chain-of-thought explanation)","structured JSON (OpenAI response format)"],"categories":["planning-reasoning","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_5","uri":"capability://planning.reasoning.safe.agent.execution.with.nemoclaw","name":"safe agent execution with nemoclaw","description":"Provides NemoClaw, a safety-focused agent execution framework for building agentic AI systems with built-in guardrails, sandboxing, and execution monitoring. Enables controlled tool calling, function execution, and multi-step reasoning within bounded safety constraints. Integrates with NIM inference to route agent decisions through NVIDIA-optimized models while enforcing safety policies at execution boundaries.","intents":["Build agentic AI systems with safety constraints and execution monitoring","Execute multi-step agent workflows with guardrails preventing unsafe actions","Implement tool calling and function execution with safety validation","Deploy agents on edge or on-premises hardware with confidence in execution safety"],"best_for":["Teams building autonomous agents requiring safety guarantees","Enterprises deploying agents in regulated industries (finance, healthcare, critical infrastructure)","Developers implementing multi-step reasoning workflows with external tool integration"],"limitations":["NemoClaw documentation and technical specifications not provided in source material — implementation details unknown","Safety policies, guardrail types, and constraint enforcement mechanisms not documented","Integration with specific tool/function calling frameworks not specified","Performance overhead of safety monitoring not quantified","Supported agent patterns and workflow types unknown"],"requires":["NVIDIA NIM API access","NemoClaw framework (availability, installation method unknown)","Tool/function definitions (format and schema unknown)","Inference model supporting reasoning (e.g., Nemotron-3-nano-omni)"],"input_types":["agent task description (text)","tool/function definitions (schema format unknown)","execution constraints and safety policies (format unknown)"],"output_types":["agent execution result (text, structured data)","execution trace with safety validation logs (format unknown)"],"categories":["planning-reasoning","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_6","uri":"capability://image.visual.ocr.and.document.understanding.inference","name":"ocr and document understanding inference","description":"Supports optical character recognition (OCR) and document understanding tasks through NIM-optimized models, enabling extraction of text, structure, and meaning from images and scanned documents. Processes document images through inference models trained for document understanding, returning extracted text, layout information, and semantic understanding. Runs on NVIDIA GPUs with TensorRT-LLM optimization for low-latency document processing.","intents":["Extract text from scanned documents or images without external OCR services","Understand document structure and semantic content (tables, forms, sections)","Process documents on-premises or edge without sending images to cloud services","Build document processing pipelines with inference-based understanding"],"best_for":["Teams processing documents on-premises for privacy or compliance reasons","Organizations building document understanding pipelines requiring low latency","Developers integrating OCR into inference-based applications"],"limitations":["Specific OCR/document understanding models not listed in source material — unclear which models support this capability","Input image formats, resolution requirements, and size limits not documented","OCR accuracy benchmarks and language support not provided","Output format (raw text, structured JSON, layout information) not specified","Performance characteristics (latency per page, throughput) not documented"],"requires":["NVIDIA NIM API access","Document image (format and resolution requirements unknown)","Model supporting OCR/document understanding (specific model names unknown)"],"input_types":["image (document scan, photo of document — formats unknown)","structured request with image data (format unknown)"],"output_types":["extracted text (plain text or structured format unknown)","document structure information (layout, tables, sections — format unknown)","semantic understanding (entities, relationships — format unknown)"],"categories":["image-visual","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_7","uri":"capability://automation.workflow.blueprints.and.starter.templates.for.ai.applications","name":"blueprints and starter templates for ai applications","description":"Provides pre-built application templates and reference architectures (Blueprints) for common AI use cases, enabling developers to quickly scaffold applications using NIM inference. Templates include example code, configuration, and deployment instructions for patterns like chatbots, reasoning agents, document processing, and agentic workflows. Blueprints abstract common integration patterns, reducing boilerplate and accelerating time-to-deployment.","intents":["Quickly scaffold AI applications without building integration boilerplate","Learn best practices for integrating NIM inference into applications","Deploy reference implementations for common AI use cases (chatbots, agents, document processing)","Reduce development time from concept to production deployment"],"best_for":["Teams new to NVIDIA NIM wanting to understand integration patterns","Developers building common AI use cases (chatbots, reasoning agents)","Organizations prototyping AI applications quickly"],"limitations":["Available Blueprints not listed in source material — unclear which use cases are covered","Blueprint maturity, maintenance status, and update frequency unknown","Customization requirements and extensibility of templates not documented","Language support for Blueprints (Python, Node.js, etc.) not specified","Integration with specific frameworks (LangChain, LlamaIndex, etc.) not documented"],"requires":["NVIDIA NIM API access","Development environment matching Blueprint language/framework","Basic understanding of the target use case (chatbot, agent, etc.)"],"input_types":["Blueprint template (code, configuration)","Application-specific configuration (API keys, model selection, etc.)"],"output_types":["scaffolded application code (language/framework dependent)","deployment configuration (Docker, Kubernetes manifests — format unknown)","documentation and examples"],"categories":["automation-workflow","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_8","uri":"capability://automation.workflow.dgx.station.integration.and.enterprise.deployment.playbooks","name":"dgx station integration and enterprise deployment playbooks","description":"Provides integration with NVIDIA DGX Station (enterprise GPU workstation) and includes deployment playbooks for enterprise environments. Enables organizations to deploy NIM inference on existing DGX infrastructure with pre-configured networking, resource allocation, and monitoring. Playbooks document deployment patterns, performance tuning, and operational best practices for enterprise GPU clusters.","intents":["Deploy NIM inference on existing DGX Station hardware without custom configuration","Integrate inference into enterprise GPU cluster infrastructure","Follow documented best practices for production deployment on DGX hardware","Leverage existing DGX investments for AI inference workloads"],"best_for":["Enterprises with existing DGX Station infrastructure","Organizations deploying inference on owned GPU hardware","Teams requiring documented deployment patterns for enterprise environments"],"limitations":["DGX Station integration details and supported configurations not documented","Playbook contents, deployment patterns, and operational procedures not specified in source material","Performance tuning guidance and optimization recommendations unknown","Monitoring, logging, and observability integration not documented","Support for other enterprise GPU platforms (not DGX) unclear"],"requires":["NVIDIA DGX Station hardware","NVIDIA NIM container runtime","Network connectivity and infrastructure for deployment","Access to deployment playbooks (availability and format unknown)"],"input_types":["DGX Station configuration (hardware specs, network setup)","deployment playbook (documentation, configuration templates)","inference workload specifications (model, throughput requirements)"],"output_types":["deployed inference service on DGX","operational metrics and monitoring data (format unknown)","deployment documentation and runbooks"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__cap_9","uri":"capability://code.generation.editing.model.specific.performance.optimization.and.quantization","name":"model-specific performance optimization and quantization","description":"Applies model-specific TensorRT-LLM optimizations including kernel fusion, quantization (INT8, FP8, or other precision levels), and GPU memory optimization to each model in the catalog. Optimizations are pre-compiled into container images, with quantization profiles tuned for specific GPU architectures (Blackwell, Hopper, RTX Pro). Developers access optimized inference without managing quantization or kernel selection.","intents":["Achieve maximum inference throughput on NVIDIA hardware without manual optimization","Deploy models with reduced memory footprint through pre-optimized quantization","Evaluate inference performance on specific GPU architectures","Reduce inference latency through GPU-specific kernel optimizations"],"best_for":["Teams requiring maximum inference performance on NVIDIA GPUs","Organizations deploying inference at scale with throughput requirements","Developers building latency-sensitive applications"],"limitations":["Quantization levels and precision options not documented — unclear which quantization strategies are applied per model","Performance benchmarks (latency, throughput, accuracy impact) not provided","Quantization is not configurable — developers cannot select alternative precision levels","Optimization tuning process and methodology not documented","Impact of quantization on model accuracy not specified"],"requires":["NVIDIA GPU matching supported architectures (Blackwell, Hopper, RTX Pro)","NVIDIA CUDA runtime and TensorRT libraries","Container runtime for deployment"],"input_types":["model identifier (e.g., 'deepseek-v4-pro')","inference request (text prompt)"],"output_types":["optimized inference result (text completion)","performance metrics (latency, throughput — format unknown)"],"categories":["code-generation-editing","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"nvidia-nim__headline","uri":"capability://deployment.infra.ai.model.inference.microservices.platform","name":"ai model inference microservices platform","description":"NVIDIA NIM is a platform for hosting AI model inference microservices with optimized containers, compatible with OpenAI APIs, and designed for deployment on cloud, on-prem, or edge environments using NVIDIA GPUs.","intents":["best AI model inference platform","AI microservices for deployment","optimized containers for AI models","NVIDIA inference microservices for cloud","inference API for AI models","deployable AI model services"],"best_for":["high-performance AI model deployment","flexible deployment options"],"limitations":[],"requires":["NVIDIA GPUs"],"input_types":[],"output_types":[],"categories":["deployment-infra"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":56,"verified":false,"data_access_risk":"high","permissions":["NVIDIA API key (authentication mechanism unconfirmed)","OpenAI-compatible client library (Python, Node.js, etc.)","Network access to NVIDIA NIM deployment (cloud, on-prem, or edge)","NVIDIA GPU (B300, B200, H200, or RTX Pro 6000 minimum)","NVIDIA CUDA runtime and cuDNN libraries","Container runtime (Docker or Kubernetes)","Network connectivity for model serving (port configuration unknown)","Multiple NVIDIA GPUs (quantity and architecture requirements unknown)","GPU interconnect (NVLink for optimal performance, or network for distributed clusters)","Container orchestration supporting multi-GPU allocation (Kubernetes with NVIDIA device plugin, or Docker Compose)"],"failure_modes":["API compatibility is claimed but not verified in source material — exact endpoint paths and payload structures unknown","Model availability limited to NVIDIA-curated catalog; custom model deployment requirements unknown","Streaming, batch, and async response modes not documented in available material","Rate limiting, quota, and token limits per model not specified","Requires NVIDIA GPU hardware; no CPU-only inference option documented","Supported GPU models limited to Blackwell (B300, B200), Hopper (H200), and RTX Pro 6000 — compatibility with older architectures unknown","Container orchestration requirements (Kubernetes, Docker) not documented","Model quantization profiles and optimization levels not configurable in source material","Inference performance benchmarks and SLA guarantees not provided","Multi-GPU scaling configuration and setup not documented","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.7,"quality":0.9,"ecosystem":0.15000000000000002,"match_graph":0.25,"freshness":0.75,"weights":{"adoption":0.3,"quality":0.25,"ecosystem":0.15,"match_graph":0.25,"freshness":0.05}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:24.483Z","last_scraped_at":null,"last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=nvidia-nim","compare_url":"https://unfragile.ai/compare?artifact=nvidia-nim"}},"signature":"IuX61zzjJt6lkY0UKkUYJGk6aEB1yMeYTW5m+TcvpNDDcWrwLBcKwVCtBzDps54J9V37z84YzISVOa1W+qzfBQ==","signedAt":"2026-06-21T00:17:48.310Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/nvidia-nim","artifact":"https://unfragile.ai/nvidia-nim","verify":"https://unfragile.ai/api/v1/verify?slug=nvidia-nim","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}