Claude 2 (Anthropic) and Grok 3 (xAI) are both frontier AI models. Claude 2 was published in July 2023 and Grok 3 in February 2025.
Grok 3 was trained on 3.5 x 10^26 FLOP, about 90.5x the compute of Claude 2 at 3.9 x 10^24 FLOP. Training compute is the closest available proxy for how much was invested in a model, though it says nothing on its own about how well that compute was spent.
Grok 3 wins the single benchmark both were scored on. Benchmark counts are a crude scoreboard — the evaluations differ wildly in what they measure and in how saturated they are — so the per-benchmark table below matters more than the tally.
| Field | Claude 2 | Grok 3 |
|---|---|---|
| Organization | Anthropic | xAI |
| Published | Jul 11, 2023 | Feb 17, 2025 |
| Training compute | 3.9 x 10^24 FLOP | 3.5 x 10^26 FLOP |
| Parameters | — | 3T |
| Training hardware | — | NVIDIA H100 SXM5 80GB |
| Chips used | — | 80,000 |
| Training time | — | 2.2K h |
| Training cost (2023 USD) | $5M | $218M |
| Training power draw | — | 109.9 MW |
| Accessibility | API access | API access |
| Open weights | No | No |
| Country | United States | United States |
| Benchmark | Claude 2 | Grok 3 |
|---|---|---|
| Epoch Capabilities Index(ECI Score) | 119.6 | 139 |
