{"object":"list","data":[{"id":"plugsky-micro","object":"model","created":1790627136,"owned_by":"plugsky","tier":"free","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast free chat \u2014 Nemotron 3 Nano Omni 30B (~1s), tools + vision + JSON. Backup GPT-OSS 20B.","features":["chat","streaming","json-mode","tools"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-minimax","plugsky-tiny"],"name":"Plugsky Micro","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2025-03","last_updated":"2025-08","version":"v2","modality":["text"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"free","country_of_origin":"US","training_cutoff":"2024-10","license":"open weights (NVIDIA Open Model License)","best_for":"Classification, intent detection, simple chat, fast batch processing","use_cases":["classification","intent-detection","simple-chat","batch-tasks","triage"],"benchmarks":{"mmlu":70,"humaneval":78},"is_fallback":false,"recommended":false},{"id":"plugsky-lite","object":"model","created":1790627136,"owned_by":"plugsky","tier":"free","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast free chat \u2014 Nemotron 3 Nano Omni 30B (~1s), tools + JSON. Backup GPT-OSS 20B.","features":["chat","streaming","json-mode","tools"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-micro","plugsky-tiny"],"name":"Plugsky Lite","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2025-03","last_updated":"2025-08","version":"Mini 4B","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"ultra-fast","cost_tier":"free","country_of_origin":"US","training_cutoff":"2025-03","license":"NVIDIA Open Model License","best_for":"Fastest free chat with tool\/function-calling support; direct answer (~0.25s)","use_cases":["fast_chat","simple_qa","classification","tool_use","agent_loop"],"benchmarks":{"mmlu":64,"humaneval":62,"ifeval":76.9},"is_fallback":false,"recommended":false},{"id":"plugsky-plus","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Balanced everyday model \u2014 Nemotron 3 Nano Omni 30B (~1s), tools + vision. Backup GPT-OSS 20B.","features":["chat","streaming","json-mode","function-calling"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-pro","plugsky-tiny"],"name":"Plugsky Plus","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2026-02","last_updated":"2026-05","version":"V4 Flash","modality":["text"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"CN","training_cutoff":"2025-12","license":"DeepSeek License (open weights)","best_for":"Balanced general agent, fast chat at low cost","use_cases":["general-chat","agents","automation","batch-processing"],"benchmarks":{"mmlu":78,"humaneval":85},"is_fallback":false,"recommended":false},{"id":"plugsky-pro","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Default workhorse \u2014 Nemotron 3 Nano Omni 30B (~1s), tools + vision + reasoning. Backup Nemotron 3 Ultra 550B.","features":["chat","streaming","json-mode","function-calling","code"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-ultra","plugsky-max"],"name":"Plugsky Pro","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":32768,"release_date":"2026-02","last_updated":"2026-06","version":"Super 3","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2025-06","license":"Qwen License","best_for":"Balanced general purpose work, agents, coding, and long-context reasoning","use_cases":["general-chat","coding","reasoning","agents","long-context","tool-use"],"benchmarks":{"mmlu":84,"humaneval":90},"is_fallback":false,"recommended":true},{"id":"plugsky-max","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"Largest model \u2014 Nemotron 3 Ultra 550B MoE (~1.2s), deep reasoning. Backup Nemotron 3 Nano Omni.","features":["chat","streaming","json-mode","function-calling","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-ultra","plugsky-pro"],"name":"Plugsky Max","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":32768,"release_date":"2025-09","last_updated":"2025-09","version":"3-Next","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2025-06","license":"Qwen License","best_for":"Complex multi-step reasoning, deep analysis, and long-context planning","use_cases":["complex-reasoning","planning","long-context","analysis"],"benchmarks":{"mmlu":84,"humaneval":90},"is_fallback":false,"recommended":false},{"id":"plugsky-frontier","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"Top-tier capability \u2014 Nemotron 3 Ultra 550B MoE (~1.2s). Backup Nemotron 3 Nano Omni.","features":["chat","streaming","json-mode","function-calling","long-context","tools"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-max","plugsky-pro"],"name":"Plugsky Frontier","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":8192,"release_date":"2025-12","last_updated":"2026-06","version":"Large 3","modality":["text","image"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"expensive","country_of_origin":"FR","training_cutoff":"2026-03","license":"Mistral Research License (MRL)","best_for":"Frontier-tier, EU data residency, 128K context, vision","use_cases":["frontier-quality","eu-residency","vision","long-context","compliance"],"benchmarks":{"mmlu":89,"humaneval":94},"is_fallback":false,"recommended":false},{"id":"plugsky-reasoning","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Reasoning-first \u2014 Nemotron 3 Nano Omni 30B reasoning mode (~1-3s). Backup Nemotron 3 Ultra 550B.","features":["chat","streaming","reasoning","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-ultra","plugsky-pro"],"name":"Plugsky Reasoning","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":16384,"release_date":"2025-06","last_updated":"2025-12","version":"3","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"medium","cost_tier":"medium","country_of_origin":"US","training_cutoff":"2025-06","license":"NVIDIA Open Model License","best_for":"Multi-step reasoning, math, code \u2014 uses chain-of-thought","use_cases":["math","code","planning","logic","analysis"],"benchmarks":{"mmlu":84,"humaneval":91},"is_fallback":false,"recommended":false},{"id":"plugsky-ultra","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"Deepest reasoning \u2014 Nemotron 3 Ultra 550B MoE. Backup Nemotron 3 Nano Omni reasoning.","features":["chat","streaming","reasoning","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-reasoning","plugsky-pro"],"name":"Plugsky Ultra","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":32768,"release_date":"2026-03","last_updated":null,"version":"Ultra 3","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":null,"audio":null,"embedding_dim":null,"speed_tier":"medium","cost_tier":"premium","country_of_origin":"US","training_cutoff":null,"license":"NVIDIA NIM","best_for":"Complex reasoning, code, multi-step analysis","use_cases":null,"benchmarks":null,"is_fallback":false,"recommended":false},{"id":"plugsky-thinker","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Step-by-step thinker \u2014 Nemotron 3 Nano Omni reasoning mode. Backup Nemotron 3 Ultra 550B.","features":["chat","streaming","reasoning","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-reasoning","plugsky-ultra"],"name":"Plugsky Thinker","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2026-06","last_updated":null,"version":"Inkling","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":null,"audio":null,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":null,"license":"NVIDIA NIM","best_for":"Fast reasoning, classification, analysis","use_cases":null,"benchmarks":null,"is_fallback":false,"recommended":false},{"id":"plugsky-kimi","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Long-context answers \u2014 Nemotron 3 Nano Omni 30B (128K ctx). Backup Nemotron 3 Ultra 550B.","features":["chat","streaming","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-nano","plugsky-pro"],"name":"Plugsky Kimi","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2026-01","last_updated":"2026-05","version":"K2.6","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"medium","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2025-09","license":"Modified MIT","best_for":"Long-context (256K), Chinese + English, code + chat","use_cases":["long-context","chinese","code","research"],"benchmarks":{"mmlu":86.5,"humaneval":91},"is_fallback":false,"recommended":false},{"id":"plugsky-longctx","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"Long-context workhorse \u2014 Nemotron 3 Ultra 550B MoE (128K ctx). Backup Nemotron 3 Nano Omni.","features":["chat","streaming","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-nano","plugsky-pro"],"name":"Plugsky Long Context","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":8192,"release_date":"2025-12","last_updated":"2026-06","version":"Large 3","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"expensive","country_of_origin":"FR","training_cutoff":"2026-03","license":"Mistral Research License (MRL)","best_for":"Long context, EU origin, fallback model (auto-fallback target)","use_cases":["long-context","fallback","eu-residency"],"benchmarks":null,"is_fallback":true,"recommended":false},{"id":"plugsky-nano","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast long-context \u2014 Nemotron 3 Nano Omni 30B (128K ctx, ~1s). Backup Nemotron 3 Ultra 550B.","features":["chat","streaming","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-longctx","plugsky-pro"],"name":"Plugsky Nano","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":16384,"release_date":"2025-11","last_updated":"2025-12","version":"3","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"US","training_cutoff":"2025-10","license":"NVIDIA Open Model License","best_for":"NVIDIA Nemotron 3 Nano \u2014 1M context, fast MoE","use_cases":["extreme-long-context","book-analysis","codebase-qa"],"benchmarks":{"mmlu":76},"is_fallback":false,"recommended":false},{"id":"plugsky-coder","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"Coding \u2014 Nemotron 3 Ultra 550B MoE. Backup Nemotron 3 Nano Omni; peers Laguna \/ GPT-OSS.","features":["chat","streaming","tools","json-mode","code","reasoning","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-coder-ultra","plugsky-coder-fast"],"name":"Plugsky Coder","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":16384,"release_date":"2026-03","last_updated":"2026-05","version":"36B","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2026-01","license":"Apache 2.0","best_for":"Best open coding model \u2014 fast MoE, great for agents","use_cases":["coding","code-review","refactoring","agents"],"benchmarks":{"mmlu":81,"humaneval":95},"is_fallback":false,"recommended":false},{"id":"plugsky-coder-ultra","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"poolside\/laguna-xs-2.1","description":"Specialised code \u2014 Poolside Laguna XS 2.1 (~0.4s). Backup Nemotron 3 Nano Omni.","features":["chat","streaming","code","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","fallback_chain":["plugsky-coder","plugsky-coder-fast"],"name":"Plugsky Coder Ultra","family":"Poolside","underlying_model":"poolside\/laguna-xs-2.1","architecture":"dense","parameters":"2.1B","max_output":16384,"release_date":"2026-05","last_updated":null,"version":"Laguna XS 2.1","modality":["text"],"reasoning":null,"function_calling":true,"json_mode":true,"streaming":true,"vision":null,"audio":null,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":null,"license":"NVIDIA NIM","best_for":"Code generation, debugging, code review","use_cases":null,"benchmarks":null,"is_fallback":false,"recommended":false},{"id":"plugsky-coder-fast","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast coding \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode","code"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-minimax","plugsky-tiny"],"name":"Plugsky Coder Fast","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":4096,"release_date":"2024-09","last_updated":"2024-12","version":"3.2","modality":["text"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2024-08","license":"Llama 3.2 Community License","best_for":"Fast code edits, lightweight completions","use_cases":["code-edits","autocomplete","lightweight-tasks"],"benchmarks":{"mmlu":65,"humaneval":76},"is_fallback":false,"recommended":false},{"id":"plugsky-deepseek-pro","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"Reasoning + code specialist \u2014 Nemotron 3 Ultra 550B MoE. Backup Nemotron 3 Nano Omni reasoning.","features":["chat","streaming","reasoning","tools","json-mode","code"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-reasoning","plugsky-pro"],"name":"Plugsky DeepSeek Pro","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":16384,"release_date":"2026-02","last_updated":"2026-05","version":"V4 Pro","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"slow","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2025-12","license":"DeepSeek License (open weights)","best_for":"Deep reasoning + code, math problems","use_cases":["deep-reasoning","code","math","long-output"],"benchmarks":{"mmlu":87,"humaneval":93},"is_fallback":false,"recommended":false},{"id":"plugsky-deepseek-flash","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast general-purpose \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode","reasoning"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-pro","plugsky-coder-fast"],"name":"Plugsky DeepSeek Flash","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":16384,"release_date":"2026-01","last_updated":"2026-05","version":"M3","modality":["text","vision","video"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"CN","training_cutoff":"2025-12","license":"MiniMax License (proprietary)","best_for":"Fast multimodal \u2014 MiniMax-M3 (was DeepSeek V4 Flash, remapped)","use_cases":["fast-chat","vision","multimodal","agents","automation"],"benchmarks":{"mmlu":85,"humaneval":90},"is_fallback":false,"recommended":false},{"id":"plugsky-gpt-oss","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"GPT-OSS profile \u2014 served by Nemotron 3 Nano Omni (~1s) while the GPT-OSS 20B NIM worker is degraded; GPT-OSS remains the backup.","features":["chat","streaming","reasoning","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"openai\/gpt-oss-20b","fallback_chain":["plugsky-pro","plugsky-lite"],"name":"Plugsky GPT-OSS","family":"OpenAI GPT-OSS","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":16384,"release_date":"2025-08","last_updated":"2025-08","version":"120b","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"US","training_cutoff":"2024-09","license":"Apache 2.0","best_for":"Open-source GPT-OSS 120B \u2014 OpenAI's open model","use_cases":["open-source","self-hosted-compat","reasoning"],"benchmarks":{"mmlu":81,"humaneval":88},"is_fallback":false,"recommended":false},{"id":"plugsky-qwen-next","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Strong mid-tier \u2014 Nemotron 3 Nano Omni 30B. Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode","long-context","reasoning"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-pro","plugsky-max"],"name":"Plugsky Qwen Next","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":32768,"release_date":"2025-09","last_updated":"2025-09","version":"3-Next","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2025-08","license":"Apache 2.0","best_for":"Long context (256K), MoE with low active params, fast","use_cases":["long-context","efficient-moe","chat","code"],"benchmarks":{"mmlu":82,"humaneval":89},"is_fallback":false,"recommended":false},{"id":"plugsky-llama4","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-ultra-550b-a55b","description":"MoE-class \u2014 Nemotron 3 Ultra 550B MoE. Backup Nemotron 3 Nano Omni.","features":["chat","streaming","tools","json-mode","long-context","moe"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-pro","plugsky-max"],"name":"Plugsky Llama 4","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-ultra-550b-a55b","architecture":"MoE","parameters":"550B (activated: ~55B)","max_output":8192,"release_date":"2025-04","last_updated":"2025-04","version":"4","modality":["text","image"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":false,"embedding_dim":null,"speed_tier":"medium","cost_tier":"medium","country_of_origin":"US","training_cutoff":"2024-12","license":"Llama 4 Community License","best_for":"Llama 4 Maverick \u2014 efficient MoE, 128 experts","use_cases":["vision","long-context","code","multilingual"],"benchmarks":{"mmlu":84,"mmmu":60},"is_fallback":false,"recommended":false},{"id":"plugsky-minimax","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast mid-tier \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode","reasoning"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-micro","plugsky-tiny"],"name":"Plugsky MiniMax","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2026-03","last_updated":"2026-05","version":"M3","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"CN","training_cutoff":"2026-02","license":"proprietary","best_for":"Fast mid-tier chat, reasoning, tool use","use_cases":["general-chat","reasoning","tool-use","coding"],"benchmarks":{"mmlu":82},"is_fallback":false,"recommended":false},{"id":"plugsky-mistral-medium","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Real Mistral \u2014 Mistral Nemotron (~8s). Backup Nemotron 3 Nano Omni.","features":["chat","streaming","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-super-120b-a12b","fallback_chain":["plugsky-pro","plugsky-mistral-small"],"name":"Plugsky Mistral Medium","family":"NVIDIA Nemotron (Mistral-compatible tier)","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (3B active)","max_output":8192,"release_date":"2025-11","last_updated":"2025-11","version":"3.5","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"FR","training_cutoff":"2025-09","license":"Mistral Research License (MRL)","best_for":"Mistral Medium \u2014 fast 128K context","use_cases":["long-context","fast-chat","mistral-fan"],"benchmarks":{"mmlu":84},"is_fallback":false,"recommended":false},{"id":"plugsky-mistral-small","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Fast small model \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode","long-context"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-minimax","plugsky-tiny"],"name":"Plugsky Mistral Small","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2026-03","last_updated":"2026-03","version":"4","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"FR","training_cutoff":"2025-12","license":"Mistral Research License (MRL)","best_for":"Mistral Small 4 \u2014 fast 128K context","use_cases":["fast-chat","long-context","mistral-fan"],"benchmarks":{"mmlu":81},"is_fallback":false,"recommended":false},{"id":"plugsky-phi","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Compact model \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-tiny","plugsky-micro"],"name":"Plugsky Phi","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":4096,"release_date":"2025-06","last_updated":"2025-12","version":"1","modality":["text"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2025-05","license":"NVIDIA Open Model License","best_for":"Ultra-compact, very fast, low cost","use_cases":["ultra-fast","edge","classification"],"benchmarks":{"mmlu":65},"is_fallback":false,"recommended":false},{"id":"plugsky-tiny","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Small fast model \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","tools","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-micro","plugsky-minimax"],"name":"Plugsky Tiny","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2025-03","last_updated":"2025-08","version":"v2","modality":["text"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2024-10","license":"NVIDIA Open Model License","best_for":"Small, fast, low cost \u2014 for high-volume tasks","use_cases":["high-volume","classification","fast-batch"],"benchmarks":{"mmlu":70},"is_fallback":false,"recommended":false},{"id":"plugsky-nano-omni","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Omni multimodal + reasoning \u2014 Nemotron 3 Nano Omni 30B (vision + text, ~1s). Backup Nemotron 3 Ultra 550B.","features":["chat","streaming","reasoning","vision","multimodal","json-mode","tools"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-qwen-vl","plugsky-vision-fast"],"name":"Plugsky Nano Omni","family":null,"underlying_model":null,"architecture":null,"parameters":null,"max_output":null,"release_date":null,"last_updated":null,"version":null,"modality":null,"reasoning":null,"function_calling":null,"json_mode":null,"streaming":null,"vision":null,"audio":null,"embedding_dim":null,"speed_tier":null,"cost_tier":null,"country_of_origin":null,"training_cutoff":null,"license":null,"best_for":null,"use_cases":null,"benchmarks":null,"is_fallback":false,"recommended":false},{"id":"plugsky-gpt-oss-20b","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"GPT-OSS 20B profile \u2014 served by Nemotron 3 Nano Omni (~1s) while the GPT-OSS NIM worker is degraded; GPT-OSS remains the backup.","features":["chat","streaming","tools","json-mode","reasoning"],"context_window":131072,"backup_upstream":"openai\/gpt-oss-20b","fallback_chain":["plugsky-pro","plugsky-micro"],"name":"Plugsky GPT-OSS 20B","family":"OpenAI GPT-OSS","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":16384,"release_date":"2025-08","last_updated":"2025-08","version":"120b","modality":["text"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"US","training_cutoff":"2024-09","license":"Apache 2.0","best_for":"Open-source GPT-OSS 120B \u2014 OpenAI's open model","use_cases":["open-source","self-hosted-compat","reasoning"],"benchmarks":{"mmlu":81,"humaneval":88},"is_fallback":false,"recommended":false},{"id":"plugsky-lightning","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Everyday fast model \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-micro","plugsky-minimax"],"name":"Plugsky Lightning","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":8192,"release_date":"2026-08","last_updated":"2026-08","version":"3.5 Lightning","modality":["text"],"reasoning":true,"function_calling":false,"json_mode":true,"streaming":true,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":null,"license":"NVIDIA NIM","best_for":"New-gen fast chat, streaming \u2014 verified 2026-08-15","use_cases":null,"benchmarks":null,"is_fallback":false,"recommended":false},{"id":"plugsky-gemma3-nano-2b","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","description":"Tiny model \u2014 Nemotron 3 Nano Omni 30B (~1s). Backup GPT-OSS 20B.","features":["chat","streaming","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-ultra-550b-a55b","fallback_chain":["plugsky-phi","plugsky-tiny"],"name":"Plugsky Gemma 3 Nano 2B","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","architecture":"MoE","parameters":"30B (activated: ~3B)","max_output":4096,"release_date":"2025-08","last_updated":"2025-12","version":"3n","modality":["text","image","audio"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":true,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2025-06","license":"Gemma License (open weights)","best_for":"Ultra-compact, fast, multimodal (text\/image\/audio)","use_cases":["edge-deployment","multimodal","tiny"],"benchmarks":{"mmlu":60},"is_fallback":false,"recommended":false},{"id":"plugsky-vision-fast","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"meta\/llama-3.2-11b-vision-instruct","description":"Fast vision \u2014 Llama 3.2 11B Vision (~0.2s), image + text. Backup Nemotron 3 Nano Omni (omni vision).","features":["chat","streaming","vision","multimodal","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","fallback_chain":["plugsky-qwen-vl","plugsky-nano-omni"],"name":"Plugsky Vision Fast","family":"Meta Llama","underlying_model":"meta\/llama-3.2-11b-vision-instruct","architecture":"dense","parameters":"11B","max_output":4096,"release_date":"2024-09","last_updated":"2024-12","version":"3.2","modality":["text","image"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"medium","country_of_origin":"US","training_cutoff":"2024-08","license":"Llama 3.2 Community License","best_for":"Fast multimodal \u2014 image input, quick descriptions","use_cases":["vision","image-captioning","ocr","quick-multimodal"],"benchmarks":{"mmmu":50},"is_fallback":false,"recommended":false},{"id":"plugsky-qwen-vl","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"meta\/llama-3.2-11b-vision-instruct","description":"Vision-language \u2014 Llama 3.2 11B Vision (~0.2s). Backup Nemotron 3 Nano Omni (omni vision).","features":["chat","streaming","vision","multimodal","tools","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","fallback_chain":["plugsky-vision-fast","plugsky-nano-omni"],"name":"Plugsky Vision","family":"Meta Llama","underlying_model":"meta\/llama-3.2-11b-vision-instruct","architecture":"dense","parameters":"11B","max_output":16384,"release_date":"2025-09","last_updated":"2025-09","version":"3.5","modality":["text","image"],"reasoning":true,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":false,"embedding_dim":null,"speed_tier":"slow","cost_tier":"expensive","country_of_origin":"CN","training_cutoff":"2025-08","license":"Apache 2.0","best_for":"Qwen 3.5 397B \u2014 multimodal + 256K context + reasoning","use_cases":["long-context","multimodal","vision","reasoning","moe"],"benchmarks":{"mmlu":88,"mmmu":70},"is_fallback":false,"recommended":false},{"id":"plugsky-gemma-4","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"meta\/llama-3.2-11b-vision-instruct","description":"Compact multimodal \u2014 Llama 3.2 11B Vision (~0.2s). Backup Nemotron 3 Nano Omni.","features":["chat","streaming","json-mode","multimodal"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","fallback_chain":["plugsky-vision-fast","plugsky-qwen-vl"],"name":"Plugsky Gemma 4","family":"Meta Llama","underlying_model":"meta\/llama-3.2-11b-vision-instruct","architecture":"dense","parameters":"11B","max_output":4096,"release_date":"2025-08","last_updated":"2025-12","version":"3n","modality":["text","image","audio"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":true,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2025-06","license":"Gemma License (open weights)","best_for":"Fast + multimodal (text\/image\/audio) \u2014 e4b-it unavailable, using 2B","use_cases":["multimodal","fast-inference","edge-deployment"],"benchmarks":{"mmlu":60},"is_fallback":false,"recommended":false},{"id":"plugsky-gemma3-nano-4b","object":"model","created":1790627136,"owned_by":"plugsky","tier":"paid","provider":"nvidia","upstream":"meta\/llama-3.2-11b-vision-instruct","description":"Tiny multimodal \u2014 Llama 3.2 11B Vision (~0.2s). Backup Nemotron 3 Nano Omni.","features":["chat","streaming","json-mode"],"context_window":131072,"backup_upstream":"nvidia\/nemotron-3-nano-omni-30b-a3b-reasoning","fallback_chain":["plugsky-qwen-vl","plugsky-vision-fast"],"name":"Plugsky Gemma 3 Nano 4B","family":"Meta Llama","underlying_model":"meta\/llama-3.2-11b-vision-instruct","architecture":"dense","parameters":"11B","max_output":4096,"release_date":"2025-08","last_updated":"2025-12","version":"3n","modality":["text","image","audio"],"reasoning":false,"function_calling":true,"json_mode":true,"streaming":true,"vision":true,"audio":true,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2025-06","license":"Gemma License (open weights)","best_for":"Small + fast + multimodal \u2014 e4b-it unavailable, using 2B","use_cases":["multimodal","edge-deployment","fast-inference"],"benchmarks":{"mmlu":60},"is_fallback":false,"recommended":false},{"id":"plugsky-embed","object":"model","created":1790627136,"owned_by":"plugsky","tier":"special","provider":"nvidia","upstream":"nvidia\/nemotron-3-embed-1b","description":"Embeddings for RAG \/ semantic search \u2014 Nemotron 3 Embed 1B (2048-dim, ~0.1s).","features":["embeddings","rag","semantic-search"],"context_window":8192,"backup_upstream":"nvidia\/nemotron-3-embed-1b","fallback_chain":["plugsky-embed-nim","plugsky-embed-multilingual"],"name":"Plugsky Embed","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-embed-1b","architecture":"dense","parameters":"1B","max_output":4096,"release_date":"2025-01","last_updated":"2025-06","version":"v1","modality":["text"],"reasoning":false,"function_calling":false,"json_mode":false,"streaming":false,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2024-12","license":"NVIDIA Open Model License","best_for":"Default embeddings: RAG, semantic search, clustering, recommendations (4096-dim)","use_cases":["rag","semantic-search","clustering","classification"],"benchmarks":{"mtEB":64.2},"is_fallback":false,"recommended":false},{"id":"plugsky-embed-nim","object":"model","created":1790627136,"owned_by":"plugsky","tier":"special","provider":"nvidia","upstream":"nvidia\/nemotron-3-embed-1b","description":"Embeddings \u2014 Nemotron 3 Embed 1B (2048-dim, ~0.1s).","features":["embeddings","rag","semantic-search"],"context_window":8192,"backup_upstream":"nvidia\/nemotron-3-embed-1b","fallback_chain":["plugsky-embed","plugsky-embed-multilingual"],"name":"Plugsky Embed NIM","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-embed-1b","architecture":"dense","parameters":"1B","max_output":4096,"release_date":"2025-01","last_updated":"2025-06","version":"v1","modality":["text"],"reasoning":false,"function_calling":false,"json_mode":false,"streaming":false,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"US","training_cutoff":"2024-12","license":"NVIDIA Open Model License","best_for":"NVIDIA NV-Embed v1 alias for general embeddings (4096-dim)","use_cases":["rag","semantic-search","clustering"],"benchmarks":{"mtEB":64.2},"is_fallback":false,"recommended":false},{"id":"plugsky-embed-multilingual","object":"model","created":1790627136,"owned_by":"plugsky","tier":"special","provider":"nvidia","upstream":"nvidia\/nemotron-3-embed-1b","description":"Embeddings (best-effort multilingual) \u2014 Nemotron 3 Embed 1B (2048-dim, ~0.1s).","features":["embeddings","rag","semantic-search","multilingual"],"context_window":8192,"backup_upstream":"nvidia\/nemotron-3-embed-1b","fallback_chain":["plugsky-embed-nim","plugsky-embed"],"name":"Plugsky Embed Multilingual","family":"NVIDIA Nemotron","underlying_model":"nvidia\/nemotron-3-embed-1b","architecture":"dense","parameters":"1B","max_output":1024,"release_date":"2024-11","last_updated":"2025-03","version":"M3","modality":["text"],"reasoning":false,"function_calling":false,"json_mode":false,"streaming":false,"vision":false,"audio":false,"embedding_dim":null,"speed_tier":"fast","cost_tier":"cheap","country_of_origin":"CN","training_cutoff":"2024-10","license":"MIT","best_for":"Multilingual embeddings for cross-language retrieval and Arabic\/GCC content (1024-dim)","use_cases":["multilingual-rag","cross-language-retrieval","arabic-content"],"benchmarks":{"mtEB":62.1},"is_fallback":false,"recommended":false}]}