{"models":[{"id":"modal/inkling-nvfp4","type":"llm","name":"Inkling NVFP4","model_id":"opper--ep-inkling-nvfp4-server.us-west.modal.direct","capabilities":["text","tools","vision","audio","reasoning","thinking"],"speed":"medium","cost":0.238,"quality":"great","context_window":1048576,"max_output_tokens":262144,"description":"The NVFP4 quantization of Thinking Machines Lab's Inkling, a multimodal mixture-of-experts model with 975B total and 41B active parameters. It accepts text, images and audio and generates text, with configurable reasoning effort.","params":{"temperature":{"min":0,"max":2,"default":1},"max_tokens":true,"reasoning":{"supported":["none","minimal","low","medium","high","xhigh","max"],"default":"","wire":"reasoning_effort"}},"pricing":{"billing_unit":"per_mtok","input":[1.2],"output":[5],"cached_input":[0.27]},"region":"GLOBAL","thinking":true,"model":"inkling-nvfp4","family":"inkling-nvfp4","pooled":true,"access":"open","provider":{"slug":"modal","display_name":"Modal","owner_country":"US"},"service_route":{"id":"modal/shared-us-west","adapter":"modal-shared"},"maker":{"slug":"thinkingmachines","display_name":"Thinking Machines Lab","country":"US"},"compliance":{"residency":"GLOBAL","inference_location":"GLOBAL","zdr":{"logging":false,"moderation":false,"caching":true,"training":false,"subprocessors":null},"training":"opt_in","logging":"none","moderation":"none","caching":"implicit","content_storage":"ephemeral","dpa_available":true,"third_party_access":true,"transfer_mechanism":"sccs","route_id":"modal/shared-us-west"},"release":{"date":"2026-07-15","source":"vendor","url":"https://thinkingmachines.ai/model-card/inkling/"},"quantization":"nvfp4","references":[{"kind":"model_card","url":"https://huggingface.co/thinkingmachines/Inkling-NVFP4","created_at":"2026-10-07T07:52:59.951576Z","updated_at":"2026-10-07T07:52:59.951576Z"},{"kind":"model_card","url":"https://thinkingmachines.ai/model-card/inkling/","created_at":"2026-10-07T07:52:59.951576Z","updated_at":"2026-10-07T07:52:59.951576Z"}],"input_modalities":["text","image","audio"],"output_modalities":["text"]}],"total":1,"offset":0,"limit":50}