Custom Comparison
Compare 2 AI Models
Claude Sonnet 5.5 vs DeepSeek V4.1-Flash
Claude Sonnet 5.5Anthropic
DeepSeek V4.1-FlashDeepSeek
Up to 3 models. Every selection is in the address bar, so the comparison is a link you can send.
Interactive Comparison
Loading chart...
$2 / $10 per million tokens | $0.30 / $1.20 per million tokens | |
|---|---|---|
| Identity | ||
| Developer | Anthropic | DeepSeek |
| Released | Sep 2026 | Sep 2026 |
| Status | GA | GA |
| Licence | Proprietary | MIT |
| Self-hostable | No | Yes - open weights on Hugging Face |
| Cost | ||
| Blended $/1M tokensinput × 0.75 + output × 0.25 | $4.000 / 1M tokens | $0.525 / 1M tokens |
| Input price | $2.000 / 1M tokens | $0.300 / 1M tokens |
| Output price | $10.000 / 1M tokens | $1.200 / 1M tokens |
| Cached input | $0.20 per 1M tokens | $0.006 / 1M tokens (peak); $0.003 off-peak |
| Batch discount | 50% ($1 / $5 per 1M tokens) | Off-peak billing: 50% off all rates |
| Capacity | ||
| Context window | 1M tokens | 1M tokens |
| Max output | 128k tokens (up to 300k on the Message Batches API with the output-300k-2026-03-24 beta header) | 384k tokens |
| Long-context surcharge | — | None (flat rate across the 1M window) |
| Capability | ||
| Vision in | Yes | Yes |
| Audio in | No | No |
| Function calling | Yes | Yes |
| Structured output | Yes | Yes - JSON output |
| Extended reasoning | Yes - adaptive thinking, default effort high | Yes - thinking (default) and non-thinking modes |
| Web search | Yes | No |
| Code execution | Yes | No |
| Access | ||
| API | Yes | Yes |
| Chat app | Yes | Yes |
| Cloud marketplaces | Amazon Bedrock, Google Cloud, Microsoft Foundry | Not offered |
| Fine-tuning | No | Not offered (weights are open) |
| Measured quality | ||
| LMArena Elo | 1476 (checked Oct 2026) | 1475 (checked Oct 2026) |
| LMArena rank | Rank 32 | Rank 39 |
| GPQA Diamond | 96 · max effort | — |
| Elo per dollarLMArena Elo ÷ blended $/1M tokens | 369 | 2810 |
