ToolsXpo
ToolsXpo

AI & ML

AI Text Embeddings API

GET request to /v1/text-embeddings, returning json in the standard envelope.

GET/v1/text-embeddings3 credits / call Paid

Authentication

Send your API key as a bearer token. Create one free in the dashboard.

Authorization: Bearer $TOOLSXPO_KEY

Parameters

Sent as query-string parameters (or a JSON body with POST). Fields marked required must be present; everything else has a sensible default.

NameTypeDefaultDescription
modelstring—Model to use: a model id, a model name, or a tier ("economy", "balanced", "premium"). Ignored if your plan does not include it, in which case the tool's default model runs and the response reports which one answered.
aiInstructionsstring—Your own instructions, added on top of ours — tone, format, domain rules. Charged as input tokens on every call, so a long instruction raises the cost of each request.
textrequiredstring1–…—Text to convert into an embedding vector.

Example request

curl "https://api.toolsxpo.com/v1/text-embeddings?text=hello" \
  -H "Authorization: Bearer $TOOLSXPO_KEY"

Response

Every endpoint answers with the same envelope: the result in data, what the call cost and your standing in meta.

{
  "ok": true,
  "data": { /* json */ },
  "meta": { "credits": 3 }
}

meta fields (every tool)

FieldTypeDescription
meta.toolstringThe tool id that ran.
meta.authstringHow the call authenticated — `key` for an API key.
meta.creditsnumberEverything this call cost, in credits (both currencies together).
meta.baseCreditsnumberThe tool's own price — the compute-credit part of `credits`.
meta.aiCreditsnumberAI credits spent on a model step. Present only when non-zero.
meta.billedFromstringWhich balance paid: the plan allowance or the purchased credit balance.
meta.pricing.mode"flat" | "metered"Flat price per call, or metered by usage.
meta.pricing.meterUnitstringMetered tools: what is counted (e.g. `megapixel`).
meta.pricing.creditsPerUnitnumberMetered tools: credits per unit counted.
meta.quota.limitnumberYour plan's allowance for the current period.
meta.quota.usednumberAllowance used so far this period.
meta.quota.remainingnumberAllowance left this period.
meta.quota.resetsAtstring (ISO 8601)When the allowance resets.
meta.quota.creditBalancenumberPurchased credits remaining, used after the allowance.
meta.quota.scopestringWhat the quota belongs to — `account` (your workspace) for API keys.
meta.quota.planstringYour plan id.

Errors

Failures use the same envelope with ok: false and a machine-readable code.

StatusCodeMeaning
400invalid_inputA parameter failed validation. `details` lists which.
401unauthorizedMissing or invalid API key.
402quota_exceededOut of credits on this workspace.
404not_foundNo such tool.
429rate_limitedToo many requests. Retry after `Retry-After` seconds.
503service_unavailableTemporarily disabled by an administrator.

How to use the Text Embeddings API

Turn text into a numeric embedding vector for semantic search, clustering, deduplication or RAG. Send text; get the vector in data.embedding and its size in data.dimensions.

Bash
curl "https://api.toolsxpo.com/v1/text-embeddings?text=How%20do%20I%20reset%20my%20password" \
  -H "Authorization: Bearer $TOOLSXPO_KEY"

Using the vector

  • Store each vector alongside its source text in a vector store (or a plain array for small sets).
  • Compare two vectors with cosine similarity — closer to 1 means more semantically similar.
  • Always embed your query with the same model you used for the documents, or the distances are meaningless.

Tips & gotchas

  • Keep each input focused (a sentence, a paragraph, a chunk) rather than a whole document — chunk long text first.
  • data.dimensions tells you the vector size to provision in your store.
  • Paid tool. One vector per call.