Alibaba released Qwen3.8-Max-Preview on July 19, 2026, and made a claim you cannot ignore: a 2.4-trillion-parameter flagship that is, in its own words, "second only to Fable 5." What it did not release is any benchmark table backing that up — no scores, no active-parameter count, no MoE configuration. The weights are promised "soon," with no date. Early independent testing runs from lukewarm to openly hostile. This page sorts the official from the leaked from the tested, as of July 20, and we will update it as the real numbers land.
Official, leaked, tested
Official (Alibaba's own statements):
- Qwen3.8-Max-Preview is live, first through Alibaba's Token Plan subscriptions and its Qoder coding tools.
- Headline size: 2.4 trillion parameters. No active-parameter count, no expert configuration disclosed — 2.4T is a marketing number until a model card says how much of it fires per token.
- The company's positioning, verbatim from its announcement: one of the most powerful models available, "second only to Fable 5."
- Open weights are promised "soon." No date, no license named. Note the pattern: Qwen3.6-Max made a similar promise and stayed closed, hosted only on Alibaba's own platforms.
Leaked (community-circulated internal numbers, unverified):
- Coding: 12.6 points behind Claude Fable 5, with a claimed 73.2% of coding outputs matching Fable 5's.
- Cowork, a 400-task agent evaluation: ahead of Opus 4.8 Max by 5 points, Kimi K3 by 8, GLM-5.2 by 17.1, and its own predecessor Qwen3.7-Max by 44.4.
- Treat every one of these as a leak. Alibaba has published none of them, and leaked internal evals have a way of shrinking on contact with independent testing.
Tested (early hands-on, first 24 hours):
- The harshest public verdict came from an independent developer benchmark: "Not Fable 5 level. Not even remotely close" — with the marketing called misleading outright.
- Chinese-community testing (where the model has been hammered hardest) reports a stiff default style that improves with very explicit prompts, occasional identity confusion, and rendering-pipeline output that testers ranked below GLM-5.2 until prompted in detail.
- Context length is itself unsettled: 200K is the commonly reported figure, while Token Plan's own UI offers settings up to 1M. Until a model card exists, both are provisional.
The honest summary: the claim is frontier-adjacent, the evidence is not yet. That does not mean the model is weak — Qwen3.7 was genuinely good, and launch-day harness problems routinely make strong models look bad. It means nobody outside Alibaba can currently tell.
The pattern this fits
Qwen3.8 is July 2026's third trillion-class Chinese release in three weeks — after Kimi K3 (2.8T, July 16) and alongside the imminent DeepSeek V4 stable — with Grok 4.6's 2T training reportedly finishing within days. The arms race is real, and so is a newer, less charming pattern: the open-weight promise as a launch accessory. K3 shipped with weights committed "by July 27." Qwen3.8 ships with weights "soon," from a lab whose previous Max release quietly never opened. Announcing openness now buys the open-source halo; delivering it later (or not) is a separate decision. We track the difference on this page, because for developers it is the entire difference between a model you can audit and a pricing page.
What you can actually use today
Qwen3.8-Max-Preview is currently reachable mainly through Alibaba's Token Plan bundles (personal tiers from 39 yuan/month, with a credits system the community has criticized for opaque burn rates). A general OpenAI-compatible API listing, in the way developers outside China would want it, is not there yet — the same staging delay K3 is going through.
What is available now, behind one OpenAI-compatible key on Turiloop: Qwen3.7-Max and Qwen3.7-Plus — the current shipping generation this preview is supposed to dethrone — next to GLM-5.2, DeepSeek V4 and Kimi. International card, no Chinese phone number. When Qwen3.8 gets a proper API release, it joins the lineup and this page gets its real numbers; the pricing calculator will pick it up the same day.
FAQ
What is Qwen3.8? Alibaba's new flagship model family, previewed July 19, 2026 as Qwen3.8-Max-Preview: 2.4 trillion parameters, positioned by Alibaba as second only to Claude Fable 5. Official benchmarks have not been published.
Is Qwen3.8 open source? Not yet. Alibaba says open weights are coming "soon," with no date or license named. Its previous Max-tier release (Qwen3.6-Max) made similar noises and remained closed, so treat the promise as pending until weights actually land.
Is Qwen3.8 really second only to Fable 5? That is Alibaba's internal claim, published without benchmark tables. Leaked numbers support it; the first independent tests dispute it, some sharply. Wait for third-party benchmarks before believing either side.
How do I access Qwen3.8? Today, mainly via Alibaba's Token Plan subscription bundles and Qoder tools. A standard API listing has not rolled out broadly. Qwen3.7-Max and 3.7-Plus are available now through OpenAI-compatible gateways like Turiloop.
How does Qwen3.8 compare to Kimi K3? Leaked Cowork agent numbers put Qwen3.8 8 points ahead; K3 counters with published launch benchmarks and a #3 Artificial Analysis ranking. Until Qwen publishes real numbers, K3 remains the strongest independently-scored Chinese model. See the K3 breakdown.