Blogger Tests Qwen4-Max Across Coding, Long-Context and Finance — Passes All Three
Blogger Divpradeep ran the newly released Qwen4-Max (2.4T-parameter MoE, native vision, 1M-token context, positioned as a "24/7 AI teammate") through three hands-on tests — coding, long documents, and finance — and it passed all three, with its long-context capability standing out most. He suggests users quickly gauge the model using their own familiar long documents.
Confirmed
- Coding test: asked for "a playable browser game as a single HTML file from one prompt — a Three.js low-poly 3D space runner with obstacles, scoring, a game-over screen, arrow-key controls, and simple sound effects." Qwen4-Max delivered a finished product with every element in one shot.
- Long-document stress test: the blogger planted 12 contradictions in a fake 199-page annual report (covering revenue growth, margins, dividends, loan terms, even the CEO's start date) and asked the model to find all of them, list the Top 10 risks, and cite page numbers. Leveraging its 1M-token context, Qwen4-Max read the full text and identified every contradiction. The blogger also called the page-referenced analysis from a 200-page report "genuinely useful."
- Finance test: given three years of financials for a fictional company, embedded with red flags such as surging receivables, shrinking margins, and negative operating cash flow alongside continued dividends, the model was asked for ratio analysis, a 5-year DCF, a sensitivity table, and a one-page investment memo; it built the full financial model and proactively flagged those warning signs rather than just crunching numbers.
Why it matters
- The review covers three high-value scenarios — code generation, ultra-long-context retrieval, and professional financial analysis — using reproducible methods (planted traps in fake documents, red-flag financials), offering a concrete reference for judging Qwen4-Max's practicality; the 1M-token context performance in enterprise-scale long-document analysis drew particular praise from the blogger.
2026-11-18 ~ 2026-11-18 · 5 related posts
Primary sources
- [source] Qwen4-Max hands-on: 2.4T MoE with 1M context passes game, report and finance tests — Div_pradeep · 2026-11-18
- One prompt, one HTML file: Qwen4-Max builds a playable 3D space runner — Div_pradeep · 2026-11-18
- [source] Qwen4-Max finds all 12 planted contradictions in a 199-page fake annual report — Div_pradeep · 2026-11-18
- [source] Qwen4-Max builds a 5-year DCF and flags red flags in planted-bad financials — Div_pradeep · 2026-11-18
- Qwen4-Max aces coding, long-document and finance tests; long context is its edge — Div_pradeep · 2026-11-18