Moondream: Moondream 3.1

{"about":"### Moondream 3.1 and grounding in observations\n\nMoondream 3.1 is a version of the Moondream vision-language model with a mixture-of-experts architecture. The developer specifies 9 billion total parameters and approximately 2 billion active parameters per token. This version focuses on responses tied to particular objects and image regions: it matters both how a scene is described and which part of it the result refers to.\n\nThe underlying model’s official card distinguishes five skills: image question answering, captioning, object detection, pointing and segmentation. A question produces a text explanation; detection links an object to a rectangular region; a point identifies its position; segmentation marks the area it occupies. This approach is useful in systems requiring a particular output type rather than an unrestricted description of the whole image.\n\n### Available mode in Ailoot API\n\nIn Ailoot API, Moondream 3.1 is connected for text messages and text responses. Image processing and the output of visual coordinates or masks through specialized skills are outside this listing’s current interface. For a text task, supply a prepared observation description, an attribute table or output requirements and ask a question about them.\n\nFor example, if information about products or objects is already represented in text, build the request around explicit criteria: which attributes to compare, what counts as a match and how to mark a missing value. Ask it to preserve names and source values and accompany the conclusion with a short explanation based on the supplied data. This works with a textual record of observations; it does not replace inspecting the original photograph.\n\n### Context and output boundaries\n\nThis version has a stated context of 32,000 tokens in Ailoot API. A request can include instructions, examples of the required format and material to process. If the task requires a definite choice, define the allowed answers and criteria in advance: this makes the task verifiable without assumptions about information missing from the input.\n\nMoondream’s specialization in visual understanding does not mean that every numerical or logical conclusion is automatically accurate. In the official model card, the developer distinguishes detection and spatial-grounding tasks from general text-based mathematics. Choose this version with the specific input and required answer in mind; the current Ailoot API connection returns text.","alias_of":null,"aliases":[],"author":"Moondream","canonical_model":"moondream/moondream3.1-9b-a2b","ctx":null,"formats":["chat","gemini","messages","responses"],"id":"moondream/moondream3.1-9b-a2b","input":["text"],"kind":"text","max_out":null,"methods":[{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"anthropic","fast_multiplier":1,"flex_multiplier":1,"fmt":"messages","http":"POST","img_in":null,"img_out":null,"in":0.6,"input":["text"],"label":"Anthropic Messages","max_out":null,"max_tool_calls":null,"out":5.0,"output":["text"],"path":"/v1/messages","request_model_id":"moondream/moondream3.1-9b-a2b","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"openai","fast_multiplier":1,"flex_multiplier":1,"fmt":"responses","http":"POST","img_in":null,"img_out":null,"in":0.6,"input":["text"],"label":"OpenAI Responses","max_out":null,"max_tool_calls":null,"out":5.0,"output":["text"],"path":"/v1/responses","request_model_id":"moondream/moondream3.1-9b-a2b","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"openai","fast_multiplier":1,"flex_multiplier":1,"fmt":"chat","http":"POST","img_in":null,"img_out":null,"in":0.6,"input":["text"],"label":"OpenAI Chat","max_out":null,"max_tool_calls":null,"out":5.0,"output":["text"],"path":"/v1/chat/completions","request_model_id":"moondream/moondream3.1-9b-a2b","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":null,"family":"gemini","fast_multiplier":1,"flex_multiplier":1,"fmt":"gemini","http":"POST","img_in":null,"img_out":null,"in":0.6,"input":["text"],"label":"Gemini generateContent","max_out":null,"max_tool_calls":null,"out":5.0,"output":["text"],"path":"/v1beta/models/moondream/moondream3.1-9b-a2b:generateContent","request_model_id":"moondream/moondream3.1-9b-a2b","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]}],"model_aliases":[{"canonical_model_id":"moondream/moondream3.1-9b-a2b","request_model_id":"moondream/moondream3.1-9b-a2b","route_label":"Ultra"}],"name":"Moondream: Moondream 3.1","price":{"in":0.6,"in_label":null,"out":5.0,"out_label":null,"sort_in":0.6,"sort_out":5.0,"varies":false},"public_id":"moondream/moondream3.1-9b-a2b","short":"A Moondream 3.1 version focused on grounding responses in specific scene objects; text workflows are available in Ailoot API."}