# Tako Answer > Tako Answer is a paid API for AI agents from Tako at tako.com, paid per call via MPP, $0.018/call, status unknown (last checked 2026-10-11). Answers a natural language question with cited synthesis, combining Tako's curated data, charts, and web sources into a structured response ## Facts - Endpoint: POST https://tako.com/api/mpp/v1/answer?utm_source=zero.xyz - Price: $0.018/call - Payment: MPP - Status: unknown - Last checked: 2026-10-11 - Activations on Zero: 0 - Provider: Tako - Docs: https://tako.com - Website: https://tako.com - Tags: data, visualization, charts, research, search, answers, agents, reports, datasets, analytics, ai - Canonical page: https://www.zero.xyz/c/tako-answer-c4c8bfbd - Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_tv_B82ENDhKbBhycrzhum Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently. ## How to call it through Zero Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`): ```sh zero fetch --capability tako-answer-c4c8bfbd -d '' ``` Example prompt: Can you give me a cited, data-backed answer comparing Intel and Nvidia's headcount since 2013, using deep search for higher-quality results? ## When to prefer this Use this endpoint when you need a synthesized, cited answer to a natural language research question — especially when the question involves data trends, company comparisons, or factual claims that benefit from both Tako's curated knowledge graph and live web retrieval. Prefer this over /v1/search when you want a composed answer rather than a raw list of results, and over /v1/runs (async agent) when you need a synchronous response. Choose DEEP effort for higher-quality results on complex research queries, INSTANT for lowest latency with cached data, and FAST (default) for a balanced response. ## Known failure modes - Query is empty or whitespace-only — rejected with validation error (pattern: \S, minLength: 1) - Invalid effort enum value — request rejected - Web source count out of range (must be 1-20) — validation error - INSTANT effort on /v1/answer may return lower-quality cached content with no fresh retrieval - DEEP effort billed at premium tier and may have higher latency - Unavailable or ambiguous topic may return answer with limited or no data cards - Graph node IDs may not be durable across knowledge-graph rebuilds — should not be stored long-term ## How this service works Answer a question with cited synthesis ## Output A structured response containing a synthesized, cited text answer to the query, along with data cards (which may include charts, entity nodes, and metric references), web source citations, and graph node metadata. Cards link to Tako's knowledge graph and can be resolved for further detail or pinned in future searches. ## Request schema (JSON Schema) ```json { "type": "object", "title": "AnswerRequest", "required": [ "query" ], "properties": { "query": { "type": "string", "title": "Query", "pattern": "\\S", "examples": [ "Intel vs Nvidia headcount since 2013" ], "minLength": 1, "description": "Natural language search query." }, "effort": { "enum": [ "fast", "instant", "deep" ], "type": "string", "title": "SearchEffortLevel", "description": "Public effort taxonomy for search and answer. FAST is the default.\nINSTANT serves cached embeds without a new data retrieval and works in all\nenvironments. On /v1/answer it also writes the answer with the fastest\nsynthesis model, so it trades answer quality for latency. On /v1/answer FAST\nand DEEP both write the answer with a stronger model. DEEP also widens Tako\nretrieval and adds a second, higher-quality reranking pass; it's slower and\nbills at a premium tier." }, "locale": { "type": "string", "title": "Locale", "default": "en-US", "description": "BCP-47 locale tag for language and formatting.", "x-tako-published-default": "en-US on every endpoint. Converts when any endpoint infers locale from Accept-Language." }, "sources": { "type": "object", "title": "Sources", "properties": { "web": { "type": "object", "title": "WebSourceSettings", "properties": { "count": { "anyOf": [ { "type": "integer", "maximum": 20, "minimum": 1 }, { "type": "null" } ], "title": "Count", "default": null, "description": "Maximum number of web results to return, 1 to 20. Omit it and Tako uses 5 on search and 3 on answer, unless a result cap set elsewhere in the request sizes retrieval instead." }, "category": { "anyOf": [ { "enum": [ "news", "sports", "finance" ], "type": "string", "title": "WebCategory", "description": "Web search category filter. Only 'news' maps to a provider category\ntoday; 'sports' and 'finance' are accepted but have no effect yet." }, { "type": "null" } ], "default": null, "description": "Restrict web results to a category. 'news' filters to news sources. 'sports' and 'finance' are accepted but have no effect yet. Omit for a general web search." }, "highlights": { "type": "boo … (truncated) ``` ## Response schema (JSON Schema) ```json { "type": "object", "title": "AnswerResponse", "required": [ "answer", "request_id" ], "properties": { "cards": { "type": "array", "items": { "type": "object", "title": "TakoCard", "properties": { "nodes": { "anyOf": [ { "type": "array", "items": { "type": "object", "title": "TakoCardNode", "required": [ "id", "type", "name" ], "properties": { "id": { "type": "string", "title": "Id", "description": "Opaque, human-friendly public id (::::, where prefix is ent for an entity or mt for a metric). Ids are not durable across knowledge-graph rebuilds — resolve them at request time rather than storing them." }, "name": { "type": "string", "title": "Name" }, "type": { "$ref": "#/components/schemas/GraphNodeType" }, "description": { "anyOf": [ { "type": "string" }, { "type": "null" } ], "title": "Description", "default": null } }, "description": "Slim graph node for the search and answer card surface\n(TakoCard.nodes). It is deliberately narrower than GraphNode: it carries\nno `aliases` and no `subtype`. Callers that need the full node resolve\nthe id via /v1/graph/node/{id}." } }, { "type": "null" } ], "title": "Nodes", "default": null, "description": "Graph nodes (entities and metrics) behind this card. The response includes them by default. Absent for web-only cards or when node resolution was not available. Use each id with /v1/graph/node/{id} for full detail (aliases, subtype), or pass ids in sources.data.node_ids to pin these nodes in future searches." }, "title": { "anyOf": [ { "type": "string" }, { "type": "null" } ], "title": "Title", "default": null }, "card_id": { "anyOf": [ { "type": "string" }, { "type": "null" } ], "title": "Card Id", "default": null }, "content": { "anyOf": [ { "type": "object", "title": "ResultContent", "properties": { "url": { … (truncated) ``` ## More - Live health (JSON, refreshed every minute): https://www.zero.xyz/c/tako-answer-c4c8bfbd/health.json - [Zero catalog index](https://www.zero.xyz/llms.txt) - [Other services from tako.com](https://www.zero.xyz/host/tako.com/llms.txt)