Apertus 70B is the fully open flagship of the Swiss AI Initiative, a collaboration between EPFL, ETH Zurich, and CSCS that trained the model on the Alps supercomputer in Lugano and released it in September 2025. The 70 billion parameter decoder-only transformer was pretrained on 15 trillion tokens, and the project publishes far more than weights, including the training data, recipes, code, and intermediate checkpoints. Training used the Goldfish loss, an objective that intentionally limits verbatim memorization of the corpus. Multilingual reach is the standout, with more than 1,800 natively supported languages and deliberate weight on underrepresented ones, alongside tool calling and structured output over a 64K context window. It fits teams that want a capable general model whose entire pipeline can be audited, from research and public-sector deployments to multilingual products that go beyond the usual high-resource languages.
Apertus 70B runs on 1 route through the qynio gateway. Compare residency, zero retention, and training posture at a glance โ full data-handling detail per route below.
| Provider | Region | Zero data retention | Training | Input | Output |
|---|---|---|---|---|---|
| EU | Zero data retention | No | $0.46 | $2.44 |
Each route hosting Apertus 70B has its own privacy posture, residency, and no-KYC terms. Postures are maintained by qynio with a last-verification timestamp.
Zero data retention is on by default on Pay-as-you-go โ no action required. No training on customer data. EU; deposit available.
Independent benchmark scores โ composite indices for reasoning, coding, and math, plus individual eval scores where available.
Call Apertus 70B through the qynio gateway with one API key. Let your coding agent set it up, or call it directly โ qynio is drop-in compatible with the OpenAI, Anthropic, and Google AI SDKs.
Copy this and paste it into a coding agent like Claude Code, Cursor or Codex and it'll wire up qynio for you.
import OpenAI from "openai";const client = new OpenAI({apiKey: process.env.QYN_API_KEY,baseURL: "https://api.qyn.io/v3/compat",});const completion = await client.chat.completions.create({model: "regolo/apertus-70b",messages: [{ role: "user", content: "Hello" }],});console.log(completion.choices[0].message.content);
Side-by-side on privacy, crypto rails, pricing, and benchmarks.
