SightEmbed

StableEmbeddingsPARROTSIGHT native

Multilingual semantic embeddings for search and RAG

P50 35msP95 80ms$0.02 / 1M tokens

Overview

SightEmbed produces dense vector representations tuned for cross-lingual retrieval across EN/MS/ZH corpora. It supports asymmetric retrieval (short queries against long documents) and ships with recommended index settings for pgvector, Milvus and OpenSearch.

1024-dim float vectors, L2-normalised
Query / document asymmetric modes
Batch API up to 96 inputs
Index recipes for pgvector, Milvus, OpenSearch

Capabilities

Retrieval nDCG@1090

Cross-lingual alignment87

Throughput headroom96

Cluster cohesion89

API

https://api.parrotsight.com/v1/embeddings

bash

curl -X POST https://api.parrotsight.com/v1/embeddings \
  -H "Authorization: Bearer $PS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"ps-embed-base","input":["Configure pgvector for semantic search","Index a multilingual FAQ corpus"],"encoding_format":"float"}'

Version history

v1.3.0Latest2026-06-25
Cross-lingual retrieval improvements
v1.2.0Released2026-04-02
Batch API GA

More models

View all models

ps-vision-proStable

SightVision Pro

Deep visual understanding for enterprise video assets

$1.20 / 1k calls

ps-guard-maxStable

SightGuard Max

Threat intelligence and anomaly detection at network scale

$2.80 / 1k calls