Qwen 3.8 Max Preview Is Launching Soon
· 1 day ago
This is not a leak. This is official.
@Alibaba_Qwen just dropped the real post. Qwen3.8 is launching and going open-weight soon — 2.4T parameters, framed as frontier-tier, second only to Fable 5 in their own words. And you do not have to wait for the full drop.
Qwen3.8-Max-Preview is live right now on Alibaba’s Token Plan, Qoder, and QoderWork. International token plan is up. China plan is up. Millions already saw the announcement.
While everyone else argues about Gemini 3.5 timelines, Qwen just put a Max Preview in people’s hands. That is how you steal the news cycle.
More from the feed
New model
Kimi Sold Out — Join via Invite While Plans Are Closed
Kimi sold out. New paid plans are paused. Existing members keep working. That is not a rumor — Moonshot hit GPU capacity after the Kimi K3 wave and closed new subscriptions on purpose.
People are searching kimi plan sold out, kimi soldout, kimi subscription sold out, and why is Kimi sold out. Same story every time, demand overshot compute, so new seats wait while capacity comes back in batches.
If you already pay — do nothing dramatic. Keep the seat. Do not cancel and re-buy during the pause.
If you still need in join through Moonshot’s official invite and claim membership-credit rewards while paid checkout is gated.
Disclosure, that is a The Vibe Father referral link. You can still get invite benefits (including membership credit draws). We may earn referral credit. Capacity facts still come from the official @Kimi_Moonshot post.
When batches reopen, buy the plan that matches the work — Kimi Membership (Web / App / Work) or Kimi Code Membership (coding workflows). Until then, invite for the account, API with budget caps for automation, multi-model harness for deadline work.
Full playbook (who is affected, FAQ, Membership vs Code, scam hygiene) Kimi Sold Out. Coding Plans Paused + How to Join
Primary source Moonshot capacity update
Announcement
GPT-5.6 Sol takes rank one on July 20 2026
GPT-5.6 Sol now holds rank one with a Vibe Coding Index of 64.9. The previous leader was Kimi K3.
GPT-5.6 Sol has taken the top spot in the benchmark 🚀
The refresh uses Artificial Analysis capability data through OpenRouter and Arena Code WebDev results captured July 17 2026 and July 15 2026. Preliminary source results stay marked tentative.
See the live AI coding benchmark leaderboard and read the scoring method.
Tip
How to Use Kimi K3 for Coding Today
How to use Kimi K3 for coding today 🛠️ Fifteen minutes, key, real repo task, budget alert, tests as the gate. Then put it next to Claude/Codex in a harness instead of swapping religions every hour.
