Requesty is a GDPR-compliant, EU-hosted AI gateway and LLM router (an alternative to OpenRouter). It exposes a single OpenAI-compatible endpoint in front of 400+ models from providers such as OpenAI, Anthropic, Google, DeepSeek, Meta and Mistral, with EU data residency, latency-based routing, auto-caching and per-key spend governance.
Changelog
2 versions
v2.0.0
Latest
- BREAKING: ChatCompletion renamed to SendPrompt. Its output is now `answer` and `prompt` instead of the raw `choices.message.content` path.
- BREAKING: CreateEmbedding renamed to GenerateEmbeddings. It now splits the text into chunks and returns `embeddings` ([{ index, text, vector }]) and `firstVector` instead of a single flat vector. The `input` field is now `text` and `encodingFormat` was replaced by `chunkSize` and `chunkOverlap`.
- Added GenerateEmbeddingsFromFile to generate embeddings for a stored text file.
- MakeApiCall now rejects endpoints outside the Requesty origin so the API key cannot be sent to another host.
v1.0.0
- Initial version with ChatCompletion, CreateEmbedding, ListModels and MakeApiCall components.
Actions
5
Generate Embeddings
Generates embeddings for text data using a Requesty-routed embedding model. The text is split into chunks and an embedding is returned for each chunk.
The returned embeddings is an array of the form:
TIP: use the JSONata modifier to convert the embeddings array into custom formats. For convenience, the component also returns the first vector in the embeddings array which is useful when querying vector databases to find relevant chunks.
The returned embeddings is an array of the form:
[{ "index": 0, "text": "chunk1", "vector": [1.1, 1.2, 1.3] }].TIP: use the JSONata modifier to convert the embeddings array into custom formats. For convenience, the component also returns the first vector in the embeddings array which is useful when querying vector databases to find relevant chunks.
Generate Embeddings From File
Generates embeddings for a text file using a Requesty-routed embedding model. The file is split into 1MB parts, each part is split into chunks and an embedding is returned for each chunk. The component emits one embeddings array per file part.
The returned embeddings is an array of the form:
TIP: use the JSONata modifier to convert the embeddings array into custom formats.
The returned embeddings is an array of the form:
[{ "index": 0, "text": "chunk1", "vector": [1.1, 1.2, 1.3] }].TIP: use the JSONata modifier to convert the embeddings array into custom formats.
List Models
Lists the models available through the Requesty router, including their identifiers and metadata.