Embedding API · served now
A European embedding model API: Qwen3 Embedding behind the OpenAI-compatible /v1/embeddings endpoint, served from hardware AxForge owns in Sweden — with zero prompt retention.
Specifications
| Model | Qwen3 Embedding — text embedding model, Qwen3 family |
|---|---|
| API model name | qwen3-embed |
| Vector dimensions | 1024 |
| Context window | 32,768 tokens |
| Endpoint | POST /v1/embeddings |
| Hardware | NVIDIA DGX Spark (GB10, 128 GB unified memory) — owned and operated by AxForge |
| Region | eu-se-1 · Stockholm, Sweden — pinned with x-axforge-region, echoed on the response |
| Pricing | €0.015 / 1M tokens (launch pricing) |
For quality benchmarks, see the official model card — we publish only numbers we measured ourselves.
Quickstart
$ export OPENAI_BASE_URL=https://api.axforge.ai/v1 $ curl -sS "$OPENAI_BASE_URL/embeddings" \ -H "Authorization: Bearer $AXFORGE_API_KEY" \ -H "x-axforge-region: eu-se-1" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-embed", "input":["Zero-retention embeddings, computed in Stockholm."]}'
# pip install openai from openai import OpenAI client = OpenAI( base_url="https://api.axforge.ai/v1", api_key="YOUR_AXFORGE_KEY", default_headers={"x-axforge-region": "eu-se-1"}, ) r = client.embeddings.create( model="qwen3-embed", input=["first chunk", "second chunk"], ) print(len(r.data[0].embedding)) # 1024
Data & privacy
Prompts and completions are processed in memory in Sweden — not written to disk, not logged, not retained, and never used to train anything. That includes every document you send for embedding. We keep only request metadata (token counts, timestamps, status) for billing and operations. The full policy is at axforge.ai/privacy.
FAQ
Yes. It is the standard /v1/embeddings shape, so the official
OpenAI SDKs work by changing base_url to
https://api.axforge.ai/v1. The model name is
qwen3-embed.
1024-dimensional vectors, with a 32,768-token context window per input.
In Stockholm, Sweden (region eu-se-1), on hardware AxForge
owns and operates. TLS terminates in the EU. Spain (eu-es-1) is
live for the platform; more EU regions are in deployment.
Yes — that is exactly what it is built for. The service is GDPR compliant: EU region, EU-owned hardware, and zero prompt retention — inputs are processed in memory and never persisted or used for training. See the privacy policy.
Launch pricing is €0.015 per million input tokens. For committed volume, talk to an engineer.
Yes. Pair qwen3-embed for retrieval with
Qwen3.8 27B for generation — same API,
same key, same EU region.