Qwen 4: confirmed facts, the latest teaser and the release status of Alibaba's new model

Published: November 7, 2026 · Rumors & leaks

As of 7 November 2026, Qwen 4 has not launched: there is a teaser and a confirmed-facts breakdown from Kingy AI published on 6 November. No confirmed dates, parameters or prices exist yet, so everything beyond that is expectation rather than fact.

What we know today

There is less confirmed information about Qwen 4 than people would like, which is normal for a model that has not shipped. The reference point is Kingy AI's breakdown from 6 November 2026: the outlet gathered the confirmed facts, went through the latest teaser and described the current release status. There is no public launch with announced timelines, model size or pricing as of today.

The practical takeaway is simple: Qwen 4 exists as a line of work Alibaba is talking about publicly, not as a model you can call through an API right now. Anything concrete — parameter counts, context length, training data composition, a start date — is unconfirmed. We deliberately avoid such numbers: they cannot be verified, and invented details are worse than no details at all.

The Qwen family is well known to people who work with open weights: Alibaba's models are often used as a base for local deployment and custom fine-tuning. That is exactly why a teaser for the next generation draws so much attention — people expect not only better quality but also clear distribution terms.

The teaser: what it means and what it leaves out

A teaser is a marketing format, not a technical spec. It confirms the work is ongoing and that the company is ready to discuss the generation publicly. It does not answer the questions that matter for integration. According to Kingy AI's 6 November 2026 material, the interest revolves around the teaser and the release status rather than published specifications.

• A release date does not follow from the teaser: in Kingy AI's breakdown, timing is discussed as an open question, not a scheduled date.
• Parameter count, context length and benchmark suite are not named in confirmed reports — treat any specific figures in summaries as somebody's estimate.
• Access terms and licensing are not clarified as of publication: whether weights will be open or the model will stay cloud-only is unknown.
• Token prices and limits cannot be inferred from a teaser: pricing usually appears with documentation, not with promotional material.

Release status: why "soon" is not a date

In news about large models, "soon" means exactly one thing: work is in progress. Between a teaser and an available model there are usually weeks or months — quality evaluation, documentation, access rollout, pricing policy. None of those stages is publicly confirmed for Qwen 4 as of 7 November 2026.

It also helps to separate an announcement from a launch. An announcement records intent; a launch delivers a working endpoint and documentation. Until the second happens, any "it is out" or "available to everyone" piece is premature. If official timing is confirmed, the topic moves from unconfirmed to announced and this breakdown will need a rewrite.

What to watch once details are confirmed

To avoid rereading dozens of retellings, a short checklist is enough. It will also be useful when official materials appear.

• Architecture and size: parameter count and whether it is a base or a lighter variant.
• Context: the real window length and how the model handles long documents in practice.
• License and weights: open or closed, and whether local deployment is allowed.
• Behaviour on code and multilingual tasks — that is usually where gains are expected.
• API terms: pricing, limits and access reliability.

What to use right now

While Qwen 4 is unreleased, similar work is handled by existing models. API Market's catalogue includes Kimi K3.1 Pro, built for code, agent workflows and long context — the class of tasks people usually have in mind when they watch a new generation. For technical writing and code, GLM 5.3 sits alongside it.

If the job is large documents and heavy context, Gemini 4 Argon, Google's senior model in the catalogue, is the fit. All of them are available through a single key, without a VPN, with payment in roubles. Keep an eye on the API Market model catalogue: it is updated regularly, and new releases are easiest to try where access is already set up.

How we will update this piece

Once Qwen 4 has confirmed timing, documentation or access terms, this breakdown will need rewriting: some points will age out, others will turn into facts. Until then it stays a summary of what can be stated without guessing: there is a teaser, there is public interest, there is Kingy AI's 6 November 2026 breakdown — and there is no confirmed release.

Models you can use right now

Connect in a minute

from openai import OpenAI

client = OpenAI(
    base_url="https://api-market1.shop/v1",
    api_key="sk-your-key",
)

reply = client.chat.completions.create(
    model="kimi-k3.1-pro",
    messages=[{"role": "user", "content": "Hi! Tell me about yourself."}],
)
print(reply.choices[0].message.content)

FAQ

Sources

• Qwen 4: Confirmed Facts, Latest Teaser and Release Status - Kingy AI — Kingy AI, 2026-11-06

More news

• Gemini 4 Hydrogen: what Google launched on November 12
• Gemini Decision 1: Google ships a decision model at $0.10 per 1M tokens
• Gemini 3.7 Flash: Google bets on super-fast coding and half the price
• Gemini 4 Pro leaks promise big benchmarks and a low price
• Gemini Omni 1.2 Flash: 15-second clips, edits in conversation, $0.07 per second

Русская версия