Available Models
Dieser Inhalt ist für v1.1.0. Geh zur neuesten Version, um die aktuellste Dokumentation zu bekommen.
Dieser Inhalt ist noch nicht in deiner Sprache verfügbar.
AI Foundation Services provides access to a wide range of open-source and proprietary LLMs, embedding models, vision models, and audio models — all through a single OpenAI-compatible API.
Every hosted model has an individual model card with capabilities, sovereignty rating, pricing, plan availability, and a usage example. Browse the categories in the sidebar or use the table below.
Featured Models
Section titled “Featured Models”Verfügbare Modelle (46)
Gemma 4 – Preview
GPT-OSS 120B
Qwen3-Next 80B
Qwen3-VL 30B
Qwen3 Coder 30B
Llama 3.3 70B
Mistral Small 3
Mistral Medium 3
GPT-5.2
GPT-5
GPT-5 mini
GPT-5 Codex
o4 mini
o3
o3 mini
o1 mini
Claude 4.6 Opus
Claude 4.6 Sonnet
Claude 4.5 Opus
Claude 4.5 Sonnet (>200k)
Claude 4.5 Sonnet (≤200k)
Claude 4.5 Haiku
Claude 4 Sonnet
Claude 3.7 Sonnet
Gemini 3 Pro (>200k)
Gemini 3 Pro (≤200k)
Gemini 3 Flash
Gemini 2.5 Pro (>200k)
Gemini 2.5 Pro (≤200k)
Gemini 2.5 Flash
Gemini 2.5 Flash Image
GPT-4.1
GPT-4.1 mini
GPT-4.1 nano
GPT-4o
GPT-4o mini
GPT-Image-1 (Text→Image)
GPT-Image-1 (Image→Image)
BGE-M3
Ada Text Embedding
Jina Embeddings v2 Base DE
Whisper Large v3
Whisper Large v3 Turbo
o1
Qwen 3 Coder 30B
GPT 5.2
For a complete list with pricing, see the Plans & Pricing page.
List Models via API
Section titled “List Models via API”curl "$OPENAI_BASE_URL/models" \ -H "Authorization: Bearer $OPENAI_API_KEY"from openai import OpenAI
client = OpenAI()
models = client.models.list()for model in models.data: print(model.id)import OpenAI from "openai";
const client = new OpenAI();
const models = await client.models.list();for (const model of models.data) { console.log(model.id);}Model Metadata
Section titled “Model Metadata”Each model includes metadata you can inspect:
models = client.models.list()print(models.data[0].meta_data){ "model_type": "LLM", "source_type": "OPEN SOURCE", "max_sequence_length": 128000, "hidden_size": 0, "max_output_length": 0, "deployment_region": "otc-germany", "location": "otc-germany", "license": "https://www.llama.com/llama3_3/license/", "display_name": "Meta LLama 3.3 70B", "deployment_country": "Germany, EU", "input_modalities": ["text"], "output_modalities": ["text"], "model_lifecycle_stage": "Stable", "is_externally_hosted": false}Key fields:
- model_type —
LLM,EMBEDDING,STT(audio) - max_sequence_length — Maximum input + output tokens
- hidden_size — Embedding vector dimensions (embedding models only)
- max_output_length — Maximum output tokens the model can generate
- deployment_region — Where the model is hosted (e.g.,
otc-germanyfor T-Cloud Germany)
Model Categories
Section titled “Model Categories”| Category | Examples | Use Case |
|---|---|---|
| LLM | Llama 3.3 70B, GPT 4.1, Claude Sonnet 4, Qwen 3 | Text generation, chat, reasoning |
| Embedding | jina-embeddings-v2-base-de, text-embedding-bge-m3 | Semantic search, RAG |
| Vision | Qwen3-VL-30B-A3B-Instruct-FP8, Gemini 2.5 Flash | Image analysis, multimodal chat |
| Audio/STT | whisper-large-v3, whisper-large-v3-turbo | Speech-to-text, translation |
| Image Generation | gpt-image-1 | Image creation from text prompts |
Hosting & Compliance
Section titled “Hosting & Compliance”- Open-source models are operated by T-Systems on the T-Cloud in Germany. Data is processed inside the EU/EEA, is not stored, and is not viewable by Telekom or third parties.
- Proprietary models (GPT, Claude, Gemini, Mistral Medium) are forwarded to the upstream provider (Microsoft Azure or Google Cloud), who may use sub-processors worldwide. Selecting one of these models accepts the third-party provider’s terms.
Customer data is not used for model training, regardless of which model you select.
See Enterprise Trust for the full two-category data-handling breakdown.