Does Qwen3.5 4B support zero data retention?

Not self-serve on Pay-as-you-go. Zero data retention for Qwen3.5 4B is available on request via Opper Enterprise — contact hello@opper.click.

Where is Qwen3.5 4B hosted?

Qwen3.5 4B runs on US hosting routes through the Opper gateway.

Is Qwen3.5 4B trained on prompts and outputs?

No. Every hosting route for Qwen3.5 4B is configured so customer prompts and outputs are not used to train models by default.

Qwen3.5 4B

by Alibaba

Qwen3.5 4B is a compact 4 billion parameter multimodal model that pairs vision-language capability with a hybrid architecture combining Gated DeltaNet and Mixture-of-Experts layers. It supports a 262,144 token native context window, extensible toward roughly one million tokens with YaRN scaling, and reports MMLU-Pro 79.1. Trained with early fusion on multimodal tokens, it handles image-text-to-text tasks and video understanding while staying efficient relative to larger models. It runs in thinking mode by default, generating reasoning before responses. The model suits knowledge-intensive tasks, visual reasoning, and agent-based applications with tool calling across more than 200 languages where a small multimodal model is preferred.

Open in playground Try in chat Compare

Key info

Input

Output

Features

Context window

262K

Max output

262K

Input price

$0.03 /1M

Output price

$0.15 /1M

US residency available
Zero data retention via Enterprise
No training by default

Available routes

Qwen3.5 4B runs on 1 route through the Opper gateway. Compare residency, ZDR, and training posture at a glance — full data-handling detail per route below.

Provider	Region	Zero data retention	Training	Input	Output
DeepInfra	US	Enterprise	No	$0.03	$0.15

Training posture across routes: No training on prompts by default.

Data handling per route

Each route hosting Qwen3.5 4B has its own privacy posture, residency, and GDPR terms. Postures are maintained by Opper with a last-verification timestamp.

DeepInfra — United States🇺🇸

Zero data retention is available via Opper Enterprise contract. No training on customer data. US; unknown.

Zero data retention: Available via Opper Enterprise contract.
Training: No training on customer data.
Logging: Limited debug logs
Third-party access: None disclosed
GDPR DPA: No DPA
Transfer mechanism: unknown

Benchmarks

Independent benchmark scores — composite indices for reasoning, coding, and math, plus individual eval scores where available.

Global rank#192 of 534 LLMs

TierEfficient

Output speed24 tok/s

First token0.61s

Intelligence Index20.1

Coding Index22.6

Reasoning & knowledge

GPQA Diamond

77%

Humanity's Last Exam

Long-context reasoning

56%

Coding

SciCode

16%

Agentic & tool use

Terminal-Bench Hard

18%

τ²-Bench Telecom

92%

Math & instruction following

IFBench

52%

See full leaderboard →Benchmarks via Artificial Analysis · View on AA

Get started

Call Qwen3.5 4B through the Opper gateway with one API key. Let your coding agent set it up, or call it directly — Opper is drop-in compatible with the OpenAI, Anthropic, and Google AI SDKs.

Set it up with your agent

Copy this and paste it into your coding agent — Claude Code, Cursor, Codex, and more — and it'll wire up Opper for you.

Or call it directly

1import OpenAI from "openai";
2
3const client = new OpenAI({
4  apiKey: process.env.OPPER_API_KEY,
5  baseURL: "https://api.opper.click/v3/compat",
6});
7
8const completion = await client.chat.completions.create({
9  model: "deepinfra/Qwen/Qwen3.5-4B",
10  messages: [{ role: "user", content: "Hello" }],
11});
12console.log(completion.choices[0].message.content);