{"object":"list","data":[{"id":"qwen3.8-27b-uncensored","object":"model","created":0,"owned_by":"murakumo","context_window":8192,"is_ready":false,"murakumo":{"withheld":["datacenter-undeclared","no-published-price","capacity-unmeasured","quantization-undeclared"]}},{"id":"awai-network/hokusai","object":"model","owned_by":"awai-network","endpoint":"/v1/images/generations","kind":"image","free":true,"sizes":["1024x1024","512x512","512x768","768x512","768x768"]},{"id":"ao/usagi","object":"model","owned_by":"murakumo","kind":"organism","endpoint":"/v1/chat/completions","members":["mishima","mishima"]},{"id":"ao/kame","object":"model","owned_by":"murakumo","kind":"organism","endpoint":"/v1/chat/completions","members":["mishima","mishima"]},{"id":"ao/basho","object":"model","owned_by":"murakumo","kind":"organism","endpoint":"/v1/chat/completions","members":["mishima","mishima"],"profile":true,"description":"Specialist in topos and topology: analyzes and generates ontologies by making concepts, relations, constraints, provenance, and their structure explicit."},{"id":"ao/giemon","object":"model","owned_by":"murakumo","kind":"organism","endpoint":"/v1/chat/completions","members":["mishima","mishima"],"profile":true},{"id":"ao/hanzo","object":"model","owned_by":"murakumo","kind":"organism","endpoint":"/v1/chat/completions","members":["mishima","mishima"],"profile":true},{"id":"ao/hokusai","object":"model","owned_by":"murakumo","kind":"organism","endpoint":"/v1/images/generations","members":["awai-network/hokusai"]},{"id":"mishima","object":"model","created":0,"owned_by":"murakumo","context_window":131072,"max_output_tokens":32768,"endpoint":"/v1/chat/completions","hugging_face_id":"prism-ml/Ternary-Bonsai-2-27B-gguf","quantization":"PTQ1 ternary GGUF (5.95 GB)","input_modalities":["text"],"is_ready":true,"murakumo":{"origin":"Murakumo Mishima fleet: 10 Apple M4 16 GB Metal heads plus Intel Arc Pro B70, AMD Ryzen AI Max+ 395, Jetson AGX Xavier, aiueos K16, and aiueos 6600HS-2 heads; single-slot admission over Workers VPC","aggregation_tokens_per_step":15,"max_concurrency":15,"measured_aggregate_completion_tokens_per_second":58.94,"measured_continuously_busy_head_rate_sum_tokens_per_second":110.85,"measured_at":"2026-09-20","measurement":"13 heads, 128 completion tokens x 2 sequential rounds per head: 3328 tokens / 56.47 s = 58.94 tok/s observed batch makespan; sum of continuously busy measured head rates = 110.85 tok/s","compaction":{"enabled":false,"strategy":"preserve-input-and-reject-overflow"}}},{"id":"murakumo/free","object":"model","created":0,"owned_by":"murakumo","context_window":131072,"max_output_tokens":32768,"endpoint":"/v1/chat/completions","input_modalities":["text"],"is_ready":true,"murakumo":{"route":["openrouter:qwen/qwen3.8-27b:free","openrouter:nvidia/nemotron-3-super-120b-a12b:free","openrouter:google/gemma-4-31b-it:free","fallback:mishima"],"fallback_on":"OpenRouter non-2xx (429/402 free-quota refusals start a cooldown), unreachable, or a 200 carrying an error","fallback_max_output_tokens":32768,"note":"The OpenRouter lane serves whichever free model its router picks; x-murakumo-served-model names the one that answered. Free providers may log prompts."}},{"id":"qwen3-vl-2b-instruct","object":"model","owned_by":"murakumo","input_modalities":["text","image"],"output_modalities":["text"],"context_length":4096,"max_output_tokens":512,"readiness":"ready","ready_replicas":2,"free_replicas":0,"minimum_replicas":2,"degraded":false,"source":"Qwen/Qwen3-VL-2B-Instruct-GGUF","source_revision":"52d6c8ffea26cc873ac5ad116f8631268d7eb503","execution":"owned-fleet-gpu","external_provider":false}]}