Qwen 4 Max: benchmark results
Released Qwen4 flagship with native vision,1M context and controllable low/high/xhigh reasoning (high default).2.4T backbone plus328B learned N-gram embeddings gives2.728T under the all-parameter convention;71B backbone parameters activate per token. API-only release; smaller open models are future announcements. Provider: Alibaba. Released 2026-11-16. Access: API.
Aggregate Rating 1874 ± 11, rank #30 of 10556 rated models, from 172 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA APEX-Agents | 59.14 | Pass@1 (%) | 100 |
| AA GDP.pdf - Insurance - Criterion Pass Rate | 91.75 | Criterion Pass Rate (%) | 100 |
| EnterpriseOps-Gym-AA - Calendar | 53.85 | Task Success (%) | 100 |
| LLM Stats (APEX-Agents) | 72.1 | Score (%) | 100 |
| LLM Stats (Job Bench) | 68.5 | Score (%) | 100 |
| Spring Prompt Deckbench - Deck Rating | 1585.4 | rating | 100 |
| AA GDP.pdf - Legal - Criterion Pass Rate | 88.15 | Criterion Pass Rate (%) | 99.2 |
| EnterpriseOps-Gym-AA - Teams | 52 | Task Success (%) | 99 |
| AA GDPval | 1808.69 | ELO | 98.1 |
| AA-Briefcase - Analytical Quality Elo | 1994.05 | Elo | 97.8 |
| AA-Briefcase - Presentation Elo | 1651.51 | Elo | 97.8 |
| AutomationBench-AA - Finance - Raw Objective Completion | 94.23 | Raw Objective Completion (%) | 97.7 |
Interactive version: theaggregate.ai/model?slug=qwen-4-max · How It Works · Data refreshed daily, snapshot 2026-11-17.