AI Models
The ADITO AI Runtime serves several kinds of models behind one self-hosted platform: text generation, code generation, embeddings, and speech-to-text. Text generation, embeddings, and speech-to-text use an OpenAI-compatible API; code generation uses the Anthropic Messages API and the OpenAI Responses API instead, matching what coding agents expect. All model kinds share the same infrastructure, authentication, and data-sovereignty guarantees described in AI Compliance.
Model kinds
| Kind | Purpose | Model | Page |
|---|---|---|---|
| Text Generation | Summarize, translate, rewrite, and classify text | adito-llm | Text Generation |
| Code Generation | Coding agents (Claude Code, Codex) | adito-code | Code Generation |
| Embeddings | Semantic and similarity search | adito-embed | Embeddings |
| Speech-to-Text | Transcribe audio to text | adito-stt | Speech-to-Text |
"Model" is the floating alias; configure this unless you need a fixed, unchanging version. See Model naming and versioning below for how pinning works.
Each model page follows the same structure: a model card with the technical details, example use cases, and a first how-to to get started. Model kinds with more than one pinned codename also list a "Pinned versions" table with each codename's status and sunset/retired date.
Model naming and versioning
Model identifiers such as adito-llm are ADITO's own naming — the underlying upstream model (for example Qwen) is documented on each model's page, but never appears in the API identifier itself. The same naming scheme applies across all model kinds, including future models that are not based on Qwen.
Two identifier types are available for each model:
| Goal | Identifier to configure | Sunset / retired date |
|---|---|---|
| Automatically stay on the current version | Floating alias, e.g. adito-llm | Not applicable — always points to the current GA model |
| Pin to a fixed, unchanging version | Pinned codename, e.g. adito-llm-austen | Shown once that codename is deprecated or retired |
A floating alias (e.g. adito-llm) always points to the current generally available (GA) model and automatically benefits from upgrades. This is the default used throughout this documentation.
A pinned codename (e.g. adito-llm-austen) is an immutable version. Everything that affects output — base model, quantization, sampling defaults, context window, built-in system prompt — is frozen for that codename.
Codenames are assigned in alphabetical order as new versions ship.
Each pinned version moves through three lifecycle stages: Active (GA) → Deprecated → Retired. Deprecation and retirement are communicated in advance, and the sunset/retired date is added to the tables above once it applies.
Pinning is not yet available for Code Generation (adito-code) — only the floating alias can be configured today. For the other model kinds, see each page's "Pinned versions" table for which codenames exist and their current lifecycle status.
Use the floating alias for most use cases; it follows upgrades automatically. Pin a specific codename only if your use case needs guaranteed, unchanging output over time.
See also: Text Generation | AI Compliance | AI Overview