Compare AI models
| Specifications | ||
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 1M | 1M |
| These models have no shared benchmark scores. | ||
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | — | 57% |
| Overview | ||
| Company | Qwen | SpaceXSI |
| Release date | Nov 11 2026 | Feb 17 2025 |
| Access | Closed | Closed |
| Model details | View model | View model |
Other comparisons
Qwen4-Max PreviewvsClaude Fable 5.2Grok 3vsClaude Fable 5.2Qwen4-Max PreviewvsGPT-6.1 TerraGrok 3vsGPT-6.1 TerraQwen4-Max PreviewvsGemini 4 CarbonGrok 3vsGemini 4 CarbonQwen4-Max PreviewvsMuse EmberGrok 3vsMuse EmberQwen4-Max PreviewvsDeepSeek-V4.1-Flash-1017Grok 3vsDeepSeek-V4.1-Flash-1017Qwen4-Max PreviewvsMistral Large 4Grok 3vsMistral Large 4Frequently asked questions
Qwen4-Max Preview and Grok 3 don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Grok 3 shipped 632 days before Qwen4-Max Preview, so benchmark comparisons should account for the intervening progress.
Context windows are 1M (Qwen4-Max Preview) vs 1M (Grok 3).
Direct benchmark comparisons are unavailable — Qwen4-Max Preview and Grok 3 don't publish scores on any of the same benchmarks.