{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"awesome-mistral","slug":"mistral","name":"Mistral","type":"model","url":"https://mistral.ai/en/models","page_url":"https://unfragile.ai/mistral","categories":["model-training"],"tags":[],"pricing":{"model":"unknown","free":false,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"awesome-mistral__cap_0","uri":"capability://image.visual.multimodal.text.and.image.understanding.with.256k.token.context","name":"multimodal text-and-image understanding with 256k token context","description":"Processes both text and image inputs simultaneously within a 256k token context window, enabling analysis of documents with embedded visuals, screenshots with surrounding text, and multi-page content. Mistral Large 3 uses a unified transformer architecture to fuse text and vision embeddings, allowing cross-modal reasoning where image content informs text generation and vice versa. The extended context window (256k tokens ≈ 200 pages) enables processing of entire documents without chunking.","intents":["I need to analyze a PDF with charts, tables, and text all together without splitting it up","I want to describe what's happening in a screenshot and get code suggestions based on the UI","I need to extract structured data from multi-page documents with mixed text and images"],"best_for":["Document analysis teams processing mixed-media content","Enterprise users handling PDFs with embedded visuals","Developers building document intelligence applications"],"limitations":["Image input format support not specified (JPEG, PNG, WebP, etc. unknown)","No documented maximum image resolution or quantity per request","Vision capabilities not benchmarked against specialized vision models like GPT-4V","Context window shared between text and images — large images consume more tokens"],"requires":["API access to Mistral Large 3 model","Image files in supported formats (format list not documented)","Sufficient API quota for 256k token requests"],"input_types":["text","image (format unspecified)"],"output_types":["text","structured analysis"],"categories":["image-visual","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_1","uri":"capability://planning.reasoning.transparent.chain.of.thought.reasoning.with.explicit.reasoning.tokens","name":"transparent chain-of-thought reasoning with explicit reasoning tokens","description":"Magistral model exposes its internal reasoning process through explicit reasoning tokens that show step-by-step problem decomposition before generating final answers. This architecture allocates a portion of the token budget to internal reasoning (similar to OpenAI's o1 approach) rather than direct output generation, enabling verification of reasoning quality and debugging of incorrect conclusions. Users can inspect the reasoning trace to understand how the model arrived at its answer.","intents":["I need to verify the model's reasoning process for complex decisions, not just trust the final answer","I want to debug why the model made an incorrect conclusion by examining its intermediate steps","I need to use the model for high-stakes reasoning where explainability is required"],"best_for":["Teams building AI systems for regulated industries (finance, healthcare, legal)","Researchers studying model reasoning and failure modes","Developers building AI agents that need to justify decisions"],"limitations":["Reasoning tokens consume part of the output token budget, reducing final answer length","No documented benchmark comparing reasoning quality to non-reasoning models","Reasoning process is model-internal; no guarantee reasoning matches human logic","Likely higher latency than non-reasoning models due to extended token generation","Pricing model for reasoning tokens not documented (may cost more than standard tokens)"],"requires":["API access to Magistral model","Support for parsing reasoning token output format (format not specified)"],"input_types":["text"],"output_types":["text (reasoning tokens + final answer)"],"categories":["planning-reasoning","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_10","uri":"capability://tool.use.integration.mistral.studio.low.code.agent.and.application.builder","name":"mistral studio: low-code agent and application builder","description":"Mistral Studio is a web-based IDE for building AI agents and applications without writing code. Users define agent behavior through a visual interface, connect tools/APIs, and deploy agents directly. The platform abstracts away prompt engineering and API integration complexity, enabling non-technical users to build functional AI applications. Agents built in Studio can be deployed as APIs or embedded in applications.","intents":["I want to build an AI agent without writing code or managing prompts","I need to quickly prototype an AI application and deploy it without infrastructure setup","I want to connect my business tools (Slack, Salesforce, etc.) to an AI agent through a UI"],"best_for":["Non-technical business users building AI workflows","Product teams rapidly prototyping AI features","Organizations wanting to democratize AI application development"],"limitations":["No documentation of Studio's capabilities, UI, or supported integrations","Pricing model for Studio not documented (free tier, paid tiers unknown)","No mention of version control, collaboration features, or deployment options","Scalability and performance limits not documented","No information on custom code support or extensibility","Agent complexity limits unknown (max tools, context size, etc.)"],"requires":["Mistral Studio account (free or paid, unclear)","API key for Mistral models"],"input_types":["visual configuration"],"output_types":["deployed agent API"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_11","uri":"capability://code.generation.editing.mistral.vibe.ide.integrated.code.completion.with.real.time.suggestions","name":"mistral vibe: ide-integrated code completion with real-time suggestions","description":"Mistral Vibe is a VS Code and JetBrains IDE plugin providing real-time code completion suggestions powered by Codestral. The plugin integrates with the editor's autocomplete system, showing suggestions as the user types. Uses pay-as-you-go pricing (charged per completion request) rather than per-seat subscriptions, reducing cost for teams with variable usage. Supports multiple programming languages and includes context awareness for project-specific patterns.","intents":["I want IDE code completion that understands my codebase and suggests idiomatic patterns","I need code completion without paying per-seat subscription costs","I want to use code completion across multiple IDEs and languages with a single tool"],"best_for":["Individual developers and small teams using VS Code or JetBrains","Teams with variable code completion usage (pay-as-you-go is cheaper than subscriptions)","Developers working in multiple programming languages"],"limitations":["No documented list of supported programming languages","Context window for code completion not documented (may be limited vs API)","No mention of multi-file context awareness or codebase indexing","Latency for suggestions not documented (real-time responsiveness unknown)","No comparison to GitHub Copilot or other IDE plugins","Supported IDE versions not documented (VS Code 1.80+ mentioned but JetBrains versions unclear)"],"requires":["VS Code 1.80+ or compatible JetBrains IDE","Mistral Vibe plugin installed","Mistral API key","Internet connectivity for API calls"],"input_types":["code context (current file, cursor position)"],"output_types":["code suggestions"],"categories":["code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_12","uri":"capability://text.generation.language.le.chat.web.based.conversational.interface.with.multi.tier.pricing","name":"le chat: web-based conversational interface with multi-tier pricing","description":"Le Chat is Mistral's web-based chat interface accessible via browser, offering free and paid tiers. Free tier provides limited access to Mistral models with usage caps. Pro tier ($14.99/month) includes higher usage limits and priority access. Team tier ($24.99/month per user) adds collaboration features. Enterprise tier offers custom pricing and dedicated support. Web interface integrates web search, file uploads, and conversation history without requiring API integration.","intents":["I want to try Mistral models without writing code or managing API keys","I need a team collaboration tool for AI-assisted work with shared conversations","I want web search integration to answer questions about current events"],"best_for":["Non-technical users exploring Mistral models","Teams collaborating on AI-assisted tasks","Users wanting web search integration without API complexity"],"limitations":["Free tier usage limits not documented (requests per day, tokens per month, etc.)","No mention of conversation export or API access from Le Chat","File upload formats and size limits not documented","No mention of custom instructions or system prompts","Conversation history retention policy not documented","No offline access or local deployment option"],"requires":["Web browser with internet connectivity","Mistral account (free or paid)"],"input_types":["text","files (format unspecified)"],"output_types":["text"],"categories":["text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_13","uri":"capability://text.generation.language.benchmark.verified.performance.81.mmlu.on.mistral.small.3","name":"benchmark-verified performance: 81% mmlu on mistral small 3","description":"Mistral Small 3 achieves 81% accuracy on the MMLU (Massive Multitask Language Understanding) benchmark, a standard evaluation of general knowledge across 57 subjects. This benchmark result is publicly documented and verifiable, providing a concrete performance metric for model quality. MMLU score enables comparison with other models on a standardized scale (GPT-3.5 ≈ 86%, Claude 3 Haiku ≈ 75%, Llama 2 ≈ 45%).","intents":["I need to evaluate model quality using standard benchmarks before choosing a model","I want to compare Mistral Small 3 to competitors on a standardized metric","I need to justify model selection to stakeholders using published benchmark results"],"best_for":["Teams evaluating models for production deployment","Researchers comparing model capabilities","Organizations with compliance requirements for model transparency"],"limitations":["MMLU is a single benchmark; does not measure code generation, reasoning, or multimodal capabilities","No benchmarks published for other Mistral models (Large 3, Ministral, etc.)","MMLU measures general knowledge, not domain-specific performance","Benchmark results may not reflect real-world performance on specific tasks","No inference speed or latency benchmarks provided"],"requires":["Understanding of MMLU benchmark and its limitations"],"input_types":["benchmark evaluation"],"output_types":["performance metric (81% accuracy)"],"categories":["text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_14","uri":"capability://automation.workflow.inference.speed.of.150.tokens.second.on.mistral.small.3","name":"inference speed of 150 tokens/second on mistral small 3","description":"Mistral Small 3 achieves 150 tokens per second inference speed on standard hardware (hardware specification not documented). This throughput metric indicates latency for real-time applications: 150 tokens/sec ≈ 6.7ms per token, enabling sub-second responses for typical queries (100-200 tokens). Speed is likely achieved through optimized inference kernels and efficient model architecture (grouped query attention, etc.).","intents":["I need to evaluate model latency for real-time applications like chatbots or code completion","I want to estimate response time for my use case based on expected output length","I need to compare inference speed across models to choose the fastest option"],"best_for":["Teams building real-time applications with latency requirements","Developers optimizing for user experience in interactive tools","Organizations evaluating cost-per-inference for high-volume workloads"],"limitations":["Hardware specification for 150 tokens/sec benchmark not documented (GPU model, batch size, etc.)","Inference speed may vary significantly based on hardware, batch size, and quantization","No latency benchmarks for other Mistral models (Large 3, Ministral, etc.)","Speed is throughput (tokens/sec), not latency (time to first token), which matters for interactive apps","No comparison to competitors' inference speed"],"requires":["Understanding of token-per-second metric and its relationship to latency"],"input_types":["benchmark evaluation"],"output_types":["performance metric (150 tokens/sec)"],"categories":["automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_2","uri":"capability://code.generation.editing.code.generation.and.completion.with.specialized.codestral.model","name":"code generation and completion with specialized codestral model","description":"Codestral 25.01 is a code-specialized model trained with emphasis on code generation, completion, and repair across multiple programming languages. The model uses code-specific tokenization and training objectives optimized for syntax correctness and idiomatic patterns. Integrated into Mistral Vibe (CLI and IDE plugin) for in-editor code suggestions with pay-as-you-go pricing, enabling real-time code completion without subscription overhead.","intents":["I want IDE-integrated code completion that understands my project context and suggests idiomatic code","I need to generate boilerplate code, tests, or documentation from function signatures","I want to fix syntax errors and refactor code without leaving my editor"],"best_for":["Individual developers and small teams using VS Code or JetBrains IDEs","Teams wanting code generation without per-seat subscription costs","Developers working in multiple languages who need a single tool"],"limitations":["No documented list of supported programming languages (inferred: Python, JavaScript, Java, C++, etc. but unconfirmed)","No benchmark data comparing code generation quality to GitHub Copilot or Claude","IDE integration limited to Mistral Vibe plugin; no native VS Code Copilot-style integration","Context window size for Codestral not documented","No mention of multi-file context awareness or codebase indexing"],"requires":["Mistral Vibe CLI or IDE plugin installed","API key for Mistral API","VS Code 1.80+ or compatible JetBrains IDE (specific versions not documented)"],"input_types":["code","text (natural language prompts)"],"output_types":["code","code suggestions"],"categories":["code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_3","uri":"capability://text.generation.language.multilingual.text.generation.and.understanding.across.40.languages","name":"multilingual text generation and understanding across 40+ languages","description":"Mistral Large 3 and Ministral family models support multilingual input and output across 40+ languages with unified tokenization and training. The models use a shared vocabulary and transformer architecture trained on multilingual corpora, enabling code-switching (mixing languages in a single prompt) and translation-adjacent tasks without explicit translation models. No separate language selection required; language is inferred from input.","intents":["I need to generate content in multiple languages from a single model without language-specific fine-tuning","I want to process customer support tickets in mixed languages and respond in the customer's language","I need to build a chatbot that naturally handles code-switching between languages"],"best_for":["Global teams building products for international markets","Customer support platforms handling multilingual conversations","Content creation teams producing material in multiple languages"],"limitations":["No documented list of supported languages (40+ claimed but unspecified)","No benchmarks for non-English language quality (MMLU only reported for English)","Language detection is implicit; no explicit language tagging in API","No documented performance differences across language pairs","Code-switching behavior not documented (may degrade quality)"],"requires":["API access to Mistral Large 3 or Ministral models","Input text in one of the 40+ supported languages (list not provided)"],"input_types":["text"],"output_types":["text"],"categories":["text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_4","uri":"capability://data.processing.analysis.document.specific.text.extraction.and.table.handwriting.recognition","name":"document-specific text extraction and table/handwriting recognition","description":"Document AI model is specialized for extracting structured data from documents including text, tables, and handwritten content. The model uses document-specific training objectives and likely incorporates layout understanding (detecting columns, headers, footers) and optical character recognition (OCR) capabilities. Enables extraction of tabular data into structured formats and recognition of handwritten annotations without separate OCR pipelines.","intents":["I need to extract tables from PDFs and convert them to CSV or JSON without manual data entry","I want to digitize handwritten forms and extract structured data from them","I need to extract text from scanned documents while preserving layout and structure"],"best_for":["Document processing teams handling invoices, forms, and contracts","Organizations digitizing paper archives or handwritten records","Enterprises automating data extraction from unstructured documents"],"limitations":["No documented accuracy metrics for table extraction or handwriting recognition","Supported document formats not specified (PDF, images, etc.)","Maximum document size/page count not documented","No mention of layout preservation or spatial relationship understanding","Handwriting language support not documented","No comparison to specialized OCR tools like Tesseract or commercial solutions"],"requires":["API access to Document AI model","Document files in supported formats (formats not specified)"],"input_types":["document (PDF, image)"],"output_types":["text","structured data (JSON, CSV)"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_5","uri":"capability://text.generation.language.edge.optimized.inference.with.3b.14b.parameter.models","name":"edge-optimized inference with 3b-14b parameter models","description":"Ministral family (3B, 8B, 14B parameter variants) is engineered for edge deployment on resource-constrained devices including mobile phones, IoT devices, and embedded systems. Models use parameter-efficient architectures (likely including techniques like grouped query attention, knowledge distillation, or pruning) to maintain capability while reducing memory footprint and inference latency. Enables on-device inference without cloud connectivity, reducing latency to <100ms and eliminating API costs.","intents":["I need to run an LLM on a mobile app without sending data to the cloud","I want to deploy AI to IoT devices or robotics with limited compute and memory","I need sub-100ms inference latency for real-time applications like voice assistants"],"best_for":["Mobile app developers building on-device AI features","IoT and robotics teams with edge compute constraints","Privacy-sensitive applications requiring local inference"],"limitations":["No documented hardware requirements (GPU VRAM, CPU, RAM, storage)","No inference speed benchmarks for different hardware (mobile GPU vs CPU vs NPU)","Capability degradation vs larger models not quantified (no comparative benchmarks)","Quantization format support not documented (int8, int4, GGUF, etc.)","No mention of fine-tuning or adaptation for edge devices","Multimodal support for Ministral not confirmed (Mistral Large 3 is multimodal, but Ministral unclear)"],"requires":["Target device with sufficient RAM (amount unspecified)","Model weights in edge-compatible format (format not specified)","Inference framework supporting the model (ONNX, TensorFlow Lite, CoreML, etc. — not documented)"],"input_types":["text","possibly image (unconfirmed)"],"output_types":["text"],"categories":["text-generation-language","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_6","uri":"capability://search.retrieval.web.search.integration.with.real.time.information.retrieval","name":"web search integration with real-time information retrieval","description":"Le Chat (Mistral's web interface) integrates web search capability, enabling the model to retrieve and cite current information from the internet before generating responses. The system likely uses a search API (Google, Bing, or proprietary) to fetch relevant documents, embeds them in the context window, and generates answers with source attribution. Enables answering questions about recent events, current prices, and breaking news that are outside the model's training data cutoff.","intents":["I need answers about current events or recent news that happened after the model's training cutoff","I want the model to cite sources for factual claims so I can verify them","I need real-time information like stock prices, weather, or sports scores"],"best_for":["Users of Le Chat web interface seeking current information","Teams building chatbots that need real-time data","Researchers verifying model claims with source attribution"],"limitations":["Web search integration only available in Le Chat web interface, not in API","Search query formulation not documented (does model generate queries or use full prompt?)","No control over search sources or result filtering","Search latency not documented (likely adds 500ms-2s per request)","No documented accuracy of source attribution or hallucination rate with web results","Search results may be stale (cached) or incomplete"],"requires":["Le Chat web interface access (free or paid tier)","Internet connectivity for web search"],"input_types":["text"],"output_types":["text with source citations"],"categories":["search-retrieval","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_7","uri":"capability://planning.reasoning.agentic.reasoning.and.tool.orchestration.for.multi.step.tasks","name":"agentic reasoning and tool orchestration for multi-step tasks","description":"Mistral Large 3 includes agentic capabilities enabling the model to decompose complex tasks into subtasks, call external tools (APIs, functions), and iterate based on results. The model uses chain-of-thought reasoning to plan tool sequences and can handle tool failures by retrying or switching strategies. Enables building autonomous agents that can accomplish goals requiring multiple API calls and decision-making without explicit orchestration code.","intents":["I need to build an agent that can book flights by calling multiple APIs (search, price, book) in sequence","I want the model to autonomously decide which tools to use and in what order to solve a problem","I need an agent that can recover from tool failures and try alternative approaches"],"best_for":["Teams building autonomous agents for complex workflows","Developers creating AI assistants that interact with multiple APIs","Organizations automating multi-step business processes"],"limitations":["Tool calling mechanism not documented (function calling API format, schema validation, etc.)","No documented maximum number of tool calls per request or iteration depth","Error handling and recovery strategies not specified","No benchmarks for agent success rates or tool selection accuracy","Tool hallucination (calling non-existent tools) not addressed","No mention of cost control (agents may make expensive API calls)"],"requires":["API access to Mistral Large 3","Tool/function definitions in supported format (format not documented)","External APIs or functions to call"],"input_types":["text (task description)"],"output_types":["text (final answer) + tool calls (intermediate steps)"],"categories":["planning-reasoning","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_8","uri":"capability://automation.workflow.pay.as.you.go.api.pricing.with.per.token.billing","name":"pay-as-you-go api pricing with per-token billing","description":"Mistral offers API access with per-token billing model (input tokens and output tokens charged separately) rather than subscription-based pricing. Users pay only for tokens consumed, enabling cost-effective usage for variable workloads. Pricing structure is transparent and documented in the API dashboard, with usage tracking and spending alerts available. No minimum commitment or monthly fees required.","intents":["I want to use an LLM API without committing to a monthly subscription","I need to control costs for variable workloads that spike unpredictably","I want transparent pricing with per-request billing to allocate costs to customers"],"best_for":["Startups and small teams with variable API usage","SaaS platforms that want to pass through API costs to customers","Developers prototyping and experimenting with LLMs"],"limitations":["Actual token pricing not documented in provided materials (rates unknown)","No volume discounts or enterprise pricing mentioned","No mention of rate limiting or quota management","Billing cycle and invoice frequency not documented","No mention of cost controls or spending caps","Reasoning tokens (Magistral) pricing not documented (may cost more than standard tokens)"],"requires":["Mistral API account with payment method","API key for authentication"],"input_types":["API requests"],"output_types":["billing records, usage metrics"],"categories":["automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-mistral__cap_9","uri":"capability://automation.workflow.commercial.grade.open.weight.model.distribution.with.apache.2.0.licensing","name":"commercial-grade open-weight model distribution with apache 2.0 licensing","description":"Mistral Small 3 is distributed as an open-weight model under Apache 2.0 license, enabling free download, modification, and commercial use without licensing fees. The model weights are available in standard formats (safetensors, GGUF) for self-hosting on any infrastructure. Apache 2.0 license provides legal clarity for commercial applications and derivative works, with minimal restrictions (attribution required, no liability).","intents":["I want to download and self-host an LLM without cloud vendor lock-in","I need to fine-tune a model for my specific domain without licensing restrictions","I want to build a commercial product using an open-weight model without licensing fees"],"best_for":["Organizations prioritizing data privacy and avoiding cloud dependencies","Teams building proprietary models through fine-tuning","Developers in regulated industries (healthcare, finance) requiring model control"],"limitations":["Self-hosting requires infrastructure and operational overhead (not provided by Mistral)","No official support or SLA for self-hosted models","Model weights are large (Mistral Small 3 size not documented, likely 7-15GB)","Quantization and optimization for specific hardware is user's responsibility","No guarantee of model updates or security patches for self-hosted versions","Apache 2.0 license requires attribution in derivative works"],"requires":["GPU or CPU with sufficient VRAM (requirements not documented)","Inference framework (vLLM, Ollama, llama.cpp, etc.)","Model weights downloaded from Mistral or Hugging Face"],"input_types":["text"],"output_types":["text"],"categories":["automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":23,"verified":false,"data_access_risk":"high","permissions":["API access to Mistral Large 3 model","Image files in supported formats (format list not documented)","Sufficient API quota for 256k token requests","API access to Magistral model","Support for parsing reasoning token output format (format not specified)","Mistral Studio account (free or paid, unclear)","API key for Mistral models","VS Code 1.80+ or compatible JetBrains IDE","Mistral Vibe plugin installed","Mistral API key"],"failure_modes":["Image input format support not specified (JPEG, PNG, WebP, etc. unknown)","No documented maximum image resolution or quantity per request","Vision capabilities not benchmarked against specialized vision models like GPT-4V","Context window shared between text and images — large images consume more tokens","Reasoning tokens consume part of the output token budget, reducing final answer length","No documented benchmark comparing reasoning quality to non-reasoning models","Reasoning process is model-internal; no guarantee reasoning matches human logic","Likely higher latency than non-reasoning models due to extended token generation","Pricing model for reasoning tokens not documented (may cost more than standard tokens)","No documentation of Studio's capabilities, UI, or supported integrations","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.35,"ecosystem":0.25,"match_graph":0.25,"freshness":0.75,"weights":{"adoption":0.35,"quality":0.2,"ecosystem":0.1,"match_graph":0.3,"freshness":0.05}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-06-17T09:51:03.578Z","last_scraped_at":"2026-05-03T14:00:20.516Z","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=mistral","compare_url":"https://unfragile.ai/compare?artifact=mistral"}},"signature":"gamd9UBxWZke3Zi/r8TEgw42yTjvBkzKOpREEAN19FUKy0EKn8Ity5oy+Zkk+2+GEQi6rkMpFPEMQo2qaYwcAg==","signedAt":"2026-06-21T20:54:17.647Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/mistral","artifact":"https://unfragile.ai/mistral","verify":"https://unfragile.ai/api/v1/verify?slug=mistral","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"}}