You can deploy this model and start earning money today!

1of2/
jina-v5-omni-small-coreml-w8

Name: jina-v5-omni-small-coreml-w8
Availability: InStock

Views

No views yet

jina-v5-omni-small — Core ML bundle

On-device Core ML conversion of jinaai/jina-embeddings-v5-omni-small, produced by the Glossematics conversion pipeline. The repository root is a valid model bundle (manifest.json, formatVersion 2) consumed directly by the Glossematics Swift package — download the snapshot and point GlossModelBundle at it.

Embedding dimension: 1024 (Matryoshka: [32, 64, 128, 256, 512, 1024])
Towers: text, image, audio, video
Text buckets: [32, 64, 128, 256, 512, 1024, 2048] + ladder batch functions: b64@32, b32@64, b16@128, b8@256, b4@512
Bidirectional text tower with explicit attention mask: False
Artifacts: precompiled .mlmodelc (loads directly, no on-device compile)
Numeric recipe: int8 weights / float16 activations — embedding space jinaai/jina-embeddings-v5-omni-small:1024:w8a16

Usage (Swift, macOS 15 / iOS 18)

swift

import Glossematics

let bundle = try await GlossModelBundle(hub: "1of2/jina-v5-omni-small-coreml-w8")
let gloss = GlossOmniEmbedder(bundle: bundle)
let q = try await gloss.embed(text: "What is the capital of France?", prompt: .query)

License

The source model weights are CC-BY-NC-4.0 (© Jina AI) and this converted artifact is a derivative work distributed under the same license: non-commercial use only, with attribution. Commercial use requires a license from Jina AI. See the source model card for details.

Provenance & validation

Source: jinaai/jina-embeddings-v5-omni-small revision 41a20a1e1f56dad91e3a55d52ac6dc13007d67a5
Converter: jina_convert 0.2.0 (commit ?)
Validated against fp32 references captured from the source model (cosine parity gates; end-to-end through the public Swift API via gloss-bundle-smoke).