Alibaba's Qwen takes on Kimi K3 with open-weight Qwen 3.8, says model is "second only to Fable 5"

2026-07-19 · Source: The Decoder · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

Alibaba unveiled its Qwen 3.8 model on July 19, 2026, an open-weight model featuring 2.4 trillion parameters. The Qwen team asserts that Qwen 3.8 matches other frontier models and is "second only to Fable 5." A preview is currently accessible through Alibaba's Token Plan, Qoder, and QoderWork at 10 percent of the standard price, with open weights expected "soon." This new iteration is designed to surpass Qwen 3.7-Max, particularly in coding, full-stack development, data analysis, and office workflows. Developer Shuai Bai confirmed it is also the team's first multimodal model exceeding 1 trillion parameters, capable of processing images, videos, and documents, though no benchmark results are yet public. This release appears strategically aimed at challenging Moonshot AI's Kimi K3, which has promised open weights but currently restricts access to its chat app and API, potentially disrupting Moonshot's path to an IPO after reaching \$300 million in annual recurring revenue in June.

Key takeaway

For AI Engineers evaluating large language models, you should consider Qwen 3.8 as a strong contender, especially for complex coding, data analysis, and multimodal workflows. Its impending open weights and current 10 percent preview pricing offer a cost-effective opportunity to assess its claimed "second only to Fable 5" performance. This release signals increased competition in the frontier model space, potentially driving innovation and more accessible options for your projects.

Key insights

Alibaba's Qwen 3.8, a 2.4T parameter multimodal model, challenges Kimi K3 with open weights and claims near-Fable 5 performance.

Principles

• Open-weight models drive market disruption.
• Multimodality is key for frontier models.
• Strategic releases target competitor momentum.

In practice

• Access Qwen 3.8 preview at 10% price.
• Evaluate Qwen 3.8 for coding tasks.
• Consider Qwen 3.8 for multimodal applications.

Topics

• Qwen 3.8
• Large Language Models
• Multimodal AI
• Open-weight Models
• AI Competition
• Kimi K3

Best for: Machine Learning Engineer, Investor, CTO, AI Engineer, Director of AI/ML, Tech Journalist

Related on AIssential

• Qwen 3.8 Released : Better than Kimi K3? — Alibaba's Qwen 3.8, a 2.4 trillion parameter model, is set for an open-weight release, emphasizing deployment flexibility.
• Alibaba unveils 2.4T-parameter Qwen3.8 AI model — Alibaba's 2.4T-parameter Qwen3.8 model is launching with a new open-weight strategy, despite unverified performance claims.
• Qwen 3.8 Max IS INSANE! Second To Fable? New Open Model King? (Fully Tested) — Qwen 3.8 Max is a powerful multimodal model excelling in complex code generation and visual rendering, despite not matching top-tier benchmarks.
• Alibaba Says New AI Model Is Just Second to Anthropic’s Fable 5 — Alibaba's Qwen3.8 Max, with 2.4 trillion parameters, challenges top-tier AI models and will soon be open-weight.
• Alibaba just open-sourced a model that rivals GPT-5.2 — Alibaba's Qwen 3.5, with 17 billion active parameters, rivals top closed-source models while being open-weight.
• Is Kimi K3 Really Fable Class? — Kimi K3 shows open-weight models are rapidly closing the capability gap with frontier closed-source models.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Decoder.