Linkup /tasks batches 100 searches, fetches, and research jobs in one call at no surcharge

TL;DR: Each task is billed at its own direct-call rate - no surcharge, no discount. Batch rules: each task takes exactly the parameters of its synchronous endpoint (search: q, depth, outputType, structuredOutputSchema, includeDomains...; fetch: url, mode, renderJs, schema...; research: q, mode, reasoningDepth, outputType). Response: array of envelopes {id, type, status: pending, input, output: null, error: null}; poll GET /v1/tasks/{id} or the list endpoint with backoff (start ~5 s, cap ~30 s) until all ids are terminal.

Linkup /tasks batches 100 searches, fetches, and research jobs in one call at no surcharge

Fix

1. Each task is billed at its own direct-call rate - no surcharge, no discount. Expected: this specific failure stops.

2. The docs flag the misconception: Tasks does not make individual calls faster or cheaper - for interactive single-shot calls (chat UI, one agent step), call the synchronous endpoint directly. Expected: this specific failure stops.

Details

Batch rules: each task takes exactly the parameters of its synchronous endpoint (search: q, depth, outputType, structuredOutputSchema, includeDomains...; fetch: url, mode, renderJs, schema...; research: q, mode, reasoningDepth, outputType). Response: array of envelopes {id, type, status: pending, input, output: null, error: null}; poll GET /v1/tasks/{id} or the list endpoint with backoff (start ~5 s, cap ~30 s) until all ids are terminal. Split workloads over 100 items across submissions. Search/fetch tasks finish in seconds, research in minutes. On failed: error is a string, no charge. Context: Official docs (Linkup linkup-for-agents knowledge guide, Tasks Endpoint): /v1/tasks is an asynchronous batch wrapper around Search, Fetch, and Research. One POST accepts up to 100 {type, input} objects in any mix of search, fetch, and research.

When to use

You hit exactly this: Linkup /tasks batches 100 searches, fetches, and research jobs in one call at no surcharge in Linkup.

When not to use

A different error, or the same symptom in a different tool. This page only covers the failure above.

Compatibility

Linkup (versions mentioned here: v1).

Official docs (Linkup linkup-for-agents knowledge guide, Tasks Endpoint): /v1/tasks is an asynchronous batch wrapper around Search, Fetch, and Research. One POST accepts up to 100 {type, input} objects in any mix of search, fetch, and research. The docs flag the misconception: Tasks does not make individual calls faster or cheaper - for interactive single-shot calls (chat UI, one agent step), call the synchronous endpoint directly. Verified answer: Batch rules: each task takes exactly the parameters of its synchronous endpoint (search: q, depth, outputType, structuredOutputSchema, includeDomains...; fetch: url, mode, renderJs, schema...; research: q, mode, reasoningDepth, outputType). Each task is billed at its own direct-call rate - no surcharge, no discount. Response: array of envelopes {id, type, status: pending, input, output: null, error: null}; poll GET /v1/tasks/{id} or the list endpoint with backoff (start ~5 s, cap ~30 s) until all ids are terminal. Split workloads over 100 items across submissions. Search/fetch tasks finish in seconds, research in minutes. On failed: error is a string, no charge. Source: https://github.com/linkupplatform/linkup-for-agents/blob/HEAD/knowledge/LINKUPSPECIALIZEDENDPOINTS.md#tasks-endpoint-v1tasks