# Desert Ant Labs > The fastest models for your task, running on-device. Every model is free up to 100k monthly active devices per SDK. Unlimited inference per user. ## Models (available) - [Redact: PII redaction](https://desertant.com/models/redact/): On-device redaction that strips personal data as you type, in 27 languages, so it never reaches your servers or logs. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/redact), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/redact/). - [Emo: Emoji suggestions](https://desertant.com/models/emo/): Type a task or a message and get the emoji that fit, in 22 languages, on the device. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/emo), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/emo/). - [Gist: Content topic tagging](https://desertant.com/models/gist/): On-device topic tagging that reads a title or a post and tells you its topics, in 101 languages, with no server call. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/gist), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/gist/). - [Shapes: Shape recognition](https://desertant.com/models/shapes/): On-device shape recognition that turns one hand-drawn stroke into a rectangle, ellipse, triangle, star or line in under 10ms. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/shapes), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/shapes/). - [Clear: Speech enhancement](https://desertant.com/models/clear/): On-device speech enhancement that gives your recording the warm, close-miked sound of a podcast studio, processed 100% on-device. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/clear), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/clear/). - [Uhm: Filler-word detection](https://desertant.com/models/uhm/): On-device filler detection that marks every um, uh and hmm to within 20ms, an hour of audio in 12s. Platforms: Swift. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/uhm), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/uhm/). - [Tongue: Language identification](https://desertant.com/models/tongue/): On-device language identification that names the language of a few words, across 84 languages, in tens of microseconds. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/tongue), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/tongue/). - [Align: Word timestamps](https://desertant.com/models/align/): Word timestamps on device, for any transcriber. Cut, caption, and highlight on the word, at half the error, from a 0.7MB model. Platforms: Swift. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/align), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/align/). - [Clips: Clip selection](https://desertant.com/models/clips/): Clip selection that turns a long recording into ranked shorts and highlights from a transcript. On-device on iPhone or Mac. Platforms: Swift. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/clips), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/clips/). - [Ear: Spoken language detection](https://desertant.com/models/ear/): On-device spoken language identification across 99 languages, from a short stretch of audio, before transcription starts. Platforms: Swift, Kotlin, JS. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/ear), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/ear/). - [Title: Titles and descriptions](https://desertant.com/models/title/): On-device titles and descriptions: a short factual title and a one or two sentence description for any passage, on Apple silicon. Platforms: Swift. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/title), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/title/). - [Voz: Speech recognition](https://desertant.com/models/voz/): Transcribe 10 minutes of audio in 2s on an iPhone, 4.7x faster than Whisper, accurate word timestamps. Platforms: Swift. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/voz), [SDK](https://github.com/Desert-Ant-Labs/desert-ant-core), [Docs](https://desertant.com/docs/voz/). ## Models (closed beta) - [Moderator: Content moderation](https://desertant.com/models/moderator/): On-device NSFW detection that flags nudity before upload, trained only on licensed and synthetic data, safe to ship commercially. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/moderator). - [Schemer: Structured extraction](https://desertant.com/models/schemer/): Extract typed fields from any text against your own schema, on the device, with a field left empty where the text says nothing. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/schemer). - [Eye: Image and video scoring](https://desertant.com/models/eye/): On-device frame scoring that tells you which shot to keep from a burst or a clip, what is in the shot, and which shots repeat. - [Who: Who said what](https://desertant.com/models/who/): On-device speaker labeling that turns a recording into per-person turns with face boxes and timestamps, ready to cut or caption. - [Face: Face matching](https://desertant.com/models/face/): On-device face matching that finds the same person across your photo library or through a video, with no image leaving the device. - [Toxic: Hate speech triage](https://desertant.com/models/toxic/): On-device moderation that flags hateful, abusive and threatening text in 23 European languages, before the message is sent. Sources: [Hugging Face](https://huggingface.co/desert-ant-labs/toxic). ## Platforms - [Swift (iOS, macOS)](https://desertant.com/swift/): Drop-in Swift packages that run small, specialized models on device, on iOS and macOS, in a few lines of code on Core ML. - [Kotlin (Android, JVM)](https://desertant.com/kotlin/): Kotlin libraries that run models on device across Android and the JVM, on LiteRT and ONNX Runtime. - [JavaScript / TypeScript (Web, Node.js)](https://desertant.com/js/): npm packages that run models in the browser and in Node, with TypeScript types included. ## Resources - [Machine-readable catalog](https://desertant.com/catalog.json): all models and platforms as JSON. - [Full content](https://desertant.com/llms-full.txt): every model overview and use case in one file. - [SDKs on GitHub](https://github.com/Desert-Ant-Labs) - [Model weights on Hugging Face](https://huggingface.co/desert-ant-labs)