
刚刚,Qwen3.8-Flash震撼发布,我们看到了Qwen4的骨架 (opens in a new tab)
就在刚刚,阿里交出了一份让人满意的答卷:发布了 Qwen3.8-Flash,同步开源 Qwen3.8-Flash-Next(两者为同一模型,前者是 API 版本名称,后者是开源版本的正式名称)。它的成绩相当亮眼。在智能体编程基准 DeepSWE 1.1 上,它拿到 58.7 分,而参数规模是它三倍的前代 Qwen3.7-Plus 只有 16.5 分;
Total 8 articles

就在刚刚,阿里交出了一份让人满意的答卷:发布了 Qwen3.8-Flash,同步开源 Qwen3.8-Flash-Next(两者为同一模型,前者是 API 版本名称,后者是开源版本的正式名称)。它的成绩相当亮眼。在智能体编程基准 DeepSWE 1.1 上,它拿到 58.7 分,而参数规模是它三倍的前代 Qwen3.7-Plus 只有 16.5 分;

APPSO uses Qwen4-Max for interactive iPhone and car disassembly demos, plus OpenSCAD, ngspice and Blender workflows, while discussing task costs. These are generated visualizations and author experiences, not physical hardware teardown validation.

Artificial Analysis reports a Qwen4-Max intelligence score of 43 and average tested-task cost of $3.27. The source also collects web, presentation, spreadsheet and 3D examples. Benchmark expenses and individual cases do not represent every business workload's cost.

智东西用动画、交互场景和 GIF 任务体验 Qwen4-Max 预览版,记录了生成耗时、控制细节与字幕错位等问题。

Qwen4 尚未发布,但一份已合并的开源推理内核代码 Pull Request(PR),却率先“透露”了架构线索。

Anthropic、谷歌、OpenAI、Kimi、DeepSeek及阿里Qwen等全球AI大厂正密集备战,在中秋国庆假期前后集中推出或预告Claude 5.2、Gemini 4 Pro、Kimi K3.1、DeepSeek V4.1 Pro、Qwen4等旗舰基模,集体围攻OpenAI此前发布的GPT-6 Astra。

The report covers Qwen4-Max's architecture, million-token context and long-horizon delivery examples. Work360 and AII show different cost reductions; complex open-ended tasks remain a weakness, while smaller models are planned for later open release.

AI Tech Review tests Qwen4-Max, Hy4, Kimi and DeepSeek using frozen market data, comparing workpapers, Excel formula compatibility and API costs. Price forecasts remain unverified; this small test does not establish trading returns or predictive reliability.