Updated March 26, 2026· Based on independent benchmark data
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) leads in intelligence with a score of 51.7 vs 48.5. For speed, Grok 4.20 Beta 0309 (Reasoning) wins at 246 tok/s vs 71 tok/s.
| Metric | Grok 4.20 Beta 0309 (Reasoning) | Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) |
|---|---|---|
| Intelligence Score | 48.5 | 51.7 |
| Coding Score | 42.2 | 50.9 |
| Math Score | N/A | N/A |
| Speed (tok/s) | 246 tok/s | 71 tok/s |
| Latency (TTFT) | 11.75s | 32.16s |
| Input Price / 1M tokens | $2.00 | $3.00 |
| Output Price / 1M tokens | $6.00 | $15 |
| Context Window |
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) outperforms Grok 4.20 Beta 0309 (Reasoning) on the intelligence index with a score of 51.7 compared to 48.5. For coding tasks, Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) has the edge with a coding score of 50.9 vs 42.2.
Grok 4.20 Beta 0309 (Reasoning) generates output significantly faster at 246 tok/s compared to Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)'s 71 tok/s, making it 3.5x faster for streaming responses. Time to first token is 11.75s for Grok 4.20 Beta 0309 (Reasoning) vs 32.16s for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort), which affects perceived responsiveness in interactive applications.
Grok 4.20 Beta 0309 (Reasoning) is more affordable at $2.00/1M input tokens ($6.00/1M output), while Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) costs $3.00/1M input ($15/1M output). For a typical workload of 100 requests per day at 2,000 tokens each, Grok 4.20 Beta 0309 (Reasoning) would cost approximately $12.00/month vs $18.00/month for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) in input costs alone.
Choose Grok 4.20 Beta 0309 (Reasoning) when you need faster output (246 tok/s). Choose Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) when you need higher intelligence (51.7), stronger coding performance (50.9).
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) scores higher on coding benchmarks (50.9 vs 42.2), making it the better choice for programming tasks.
Grok 4.20 Beta 0309 (Reasoning) is cheaper at $2.00/1M input tokens vs $3.00/1M for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort).
Grok 4.20 Beta 0309 (Reasoning) is faster, producing output at 246 tok/s compared to Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)'s 71 tok/s.
No, Grok 4.20 Beta 0309 (Reasoning) does not support image input. Neither model supports image input.
Data last synced: March 26, 2026
| N/A |
| N/A |
| Max Output Tokens | N/A | N/A |
| Input Modalities | Text | Text |
| Output Modalities | Text | Text |
| Free Tier | No | No |
It depends on your priorities. Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) scores higher on intelligence (51.7), but Grok 4.20 Beta 0309 (Reasoning) may be better for specific use cases like budget-conscious projects or speed-critical applications.