DeepSeek R1 0528 has a larger context window than Grok 3 Thinking (164K tokens vs 128K tokens).
| Spec | DeepSeek R1 0528 | Grok 3 Thinking |
|---|---|---|
| Input price | Free | — |
| Output price | Free | — |
| Context window | 164K tokens | 128K tokens |
| Released | May 2025 | Feb 2025 |
| LiveCodeBench | 73.3% | 79.4% |
Not enough votes to call it. On the specs, nothing separates them.
Grok 3 Thinking uses 25.2x more transitions
DeepSeek R1 0528 has a larger context window than Grok 3 Thinking (164K tokens vs 128K tokens).
| Spec | DeepSeek R1 0528 | Grok 3 Thinking |
|---|---|---|
| Input price | Free | — |
| Output price | Free | — |
| Context window | 164K tokens | 128K tokens |
| Released | May 2025 | Feb 2025 |
| LiveCodeBench | 73.3% | 79.4% |
Not enough votes to call it. On the specs, nothing separates them.
Grok 3 Thinking uses 25.2x more transitions