Heidelberg AI Curriculum

13tools in the catalogue

AI Bench availability is not confirmed for these catalogue entries.

Scaleway

Providers · Cloud

Commercial

Good for

Call open models through an OpenAI-compatible API in French data centers.
Rent GPU Instances when you want to run vLLM, Ollama, or another engine.
Choose hosted inference or self-managed hardware based on required control.

Watch out

Generative APIs and GPU Instances require different operating responsibilities.

Official site

Requesty

Providers · Cloud

Free tier

Good for

Route requests to many language-model providers through one compatible API.
Use analytics, policy controls, caching, and automatic failover in the gateway.
Select EU residency options when Frankfurt routing fits your deployment needs.

Watch out

The gateway adds a markup to the underlying provider model fees.

Official site

OVHcloud

Providers · Cloud

Commercial

Good for

Call open-weight models through a serverless inference API in European data centers.
Rent bare GPU instances when you want to run vLLM or Ollama yourself.
Use an OpenAI-compatible endpoint while keeping the deployment under French jurisdiction.

Watch out

Its AI-specific features may be narrower than those of dedicated inference specialists.

Official site

OpenRouter

Providers · Cloud

Free tier

Good for

Use one endpoint and key to access models from many providers.
Switch models in an n8n HTTP Request node by changing one parameter.
Compare providers and track token usage from one dashboard.

Watch out

Free models are rate-limited, and shared workflows must not expose API keys.

Official site

Microsoft Entra ID

Providers · Cloud

Free tier

Good for

Manage application sign-in through a central identity service.
Assign groups and apply conditional access policies.
Provision accounts and revoke access as memberships change.

Watch out

Applications must revalidate or revoke their own local sessions after identity changes.

Official site

IONOS AI Model Hub

Providers · Cloud

Commercial

Good for

Call open-weight models through an OpenAI-compatible inference API.
Use German-hosted infrastructure without operating model hardware.
Choose it when EU-managed hosting suits the workload's requirements.

Watch out

Its model catalog is smaller and evolving; check supported models before committing work.

Official site

Groq

Providers · Cloud

Free tier

Good for

Run supported open-weight models through Groq's inference service.
Use fast responses to iterate on prompts and batch processing.
Bring your own API key for course and personal workflows.

Watch out

Free usage has rate limits; use your own API key, never a shared project key.

Official site

EUrouter

Providers · Cloud

Free tier

Good for

Route OpenAI-compatible requests among models hosted on European servers.
Set per-key limits, spending caps, and routing preferences.
Use the cloud API when centralized model access fits your workflow.

Watch out

You must use its cloud API rather than running the router locally.

Official site

Deepgram (Nova-3)

Providers · Cloud

BYOK

Good for

Transcribe recordings in batches or caption live microphone streams.
Choose it when technical vocabulary and noisy audio need speech-to-text processing.
Its API supports both prerecorded and real-time application workflows.

Watch out

It is a developer API, not a simple website for uploading files and receiving text.

Official site

Claude API

Providers · Cloud

BYOK

Good for

Call Claude models from Python, TypeScript, or REST applications.
Choose it when you need direct control over prompts and structured outputs.
It suits scripted batch work because you supply the surrounding workflow.

Watch out

You must build application logic, memory, and any agent loop around the API.

Official site

Cerebras

Providers · Cloud

Free tier

Good for

Run open-weight models through an OpenAI-compatible inference API.
Choose it when fast responses support interactive loops or batch jobs.
Its practical boundary is open-weight model selection and smaller free-tier context.

Watch out

The free tier limits context and request rate; only open-weight models are available.

Official site

Cartesia (Sonic)

Providers · Cloud

BYOK

Good for

Stream synthesized speech for live voice assistants and interactive tutors.
Choose it when an application must start speaking with little delay.
Sonic-2 supports more than fifteen languages for developer workflows.

Watch out

It is a developer API, so someone must integrate it into an application.

Official site

Azure OpenAI

Providers · Cloud

Commercial

Good for

Call OpenAI text, audio, and vision models from applications in Azure.
Choose it when Azure integration and managed regional deployment matter.
Batch processing and provisioned throughput support different operating patterns.

Watch out

An Azure subscription is required, and available models depend on region and licensing terms.

Official site