Models
Embedding models and processing modes available through the API.
Embedding models
base
Active
Default model. Powered by mixedbread-ai/mxbai-embed-large-v1 on ONNX runtime. Suitable for semantic search, classification, clustering, and RAG.
1024
Dimensions
512
Tokens / chunk
4,096
Max tokens
Long text
Auto-split into paragraphs, embed individually, combine via token-weighted average
Query prefix
Automatic search-optimised prefix for input_type "query"
Preprocessing
Whitespace normalisation, case folding, punctuation removal
Caching
Per-paragraph Redis cache (24h TTL). Repeated text and re-embedded documents with unchanged paragraphs resolve from cache
enhanced
Coming soon
AI-augmented pipeline. A fast language model preprocesses text to extract, expand, and enrich searchable content before embedding. Query-side expansion improves recall for short or ambiguous queries.
1024
Dimensions
TBD
Latency
TBD
Accuracy gain
Document side
Keyword extraction, concept expansion, synonym injection
Query side
Short query expansion into richer semantic representations
Compatibility
Same 1024-dim output. Drop-in replacement for base
Shielded mode
shielded: true
Active
Property-preserving encryption applied to any model's output. Based on the SAP algorithm (Fuchsbauer et al. 2022). Embeddings cannot be reversed to recover input text while preserving approximate nearest-neighbour search capability.
Pipeline
01
Normalise
Unit length
02
Scale
Per-team secret factor
03
Perturb
Bounded ball noise
04
Permute
Per-team dimension shuffle
Performance
Metric
Unshielded
Shielded
Added latency
—
<1ms
Output dimensions
1024
1024
Near-identical cosine
~0.99
~0.94
Distant cosine
~0.05
~0.05
Ranking preserved
Yes
Yes
Reversible
Yes
No
Constraints
Shielded embeddings from different teams cannot be compared — each team has unique encryption parameters.
Rotating the shield master secret invalidates all existing shielded embeddings. Documents must be re-embedded.
Deterministic: the same text and team always produce the same shielded vector.