Best cell per row highlighted. Null means undisclosed — never counted as zero.
| Provider | Alibaba Cloud | xAI |
| Family / tier | QwQ · Reasoning | Grok 4 · Reasoning |
| Status | GA | GA |
| Released | 2025-03-04 | 2026-03-09 |
| Knowledge cutoff | 2024-09 | 2024-11 |
| Context window | 131K | 1M |
| Max output tokens | 33K | — |
| Input modalities | text | text, image |
| Reasoning mode | always | optional |
| Open weights | Yes | No |
| Input | $0.12 | $1.25 |
| Output | $0.18 | $2.5 |
| Cached input | — | $0.2 |
| Batch input | — | — |
| Free tier | Yes | Yes |
| Output speed (tok/s) | — | 171.4 |
| Time to first token (s) | — | 13.24 |
| Latency tier | slow | medium |
| Trains on inputs | No | Yes |
| License | Apache-2.0 | Proprietary |
| Clouds | GCP | — |
| Compliance | — | — |
| Unified score | 6.8 | 6.9 |
| Decision Maker | 7 | 7 |
| Domain Strategist | 7 | 6.5 |
| Finance Lead | 7 | 6.5 |
| Domain Practitioner | 7.5 | 7.5 |
| Power User | 6 | 6.5 |
| Skeptic | 6.5 | 6 |
| Value score | 8 | 7 |
| GPQA Diamond | 65.2 | 78.5 |
| MATH-500 | 90.6 | 87.3 |
| LiveCodeBench | 63.4 | — |
| IFEval | 83.9 | 81 |
| TAU-bench | — | 93 |
| LMArena Elo | — | 1491 |
| Artificial Analysis Index | — | 49 |
Scores link to their sources. Missing cells mean the vendor hasn't disclosed a result — honesty over padding.