Head to head
OpenAI: GPT-5.6 Terra vs DeepSeek: DeepSeek V4 Flash 0423
Live prices, context windows, capabilities, and measured quality — updated continuously.
Side by side
| Metric | OpenAI: GPT-5.6 Terra Openai | DeepSeek: DeepSeek V4 Flash 0423 Deepseek |
|---|---|---|
| Input $/M tokens | $2.00 | $0.083 |
| Output $/M tokens | $12.00 | $0.165 |
| Context window | 1.1M | 1.0M |
| Max output | 128K | 384K |
| WM Score | 88 | 86 |
| Benchmark quality | 90.5 | 81 |
| Hosts serving it | 3 | 18 |
| Throughput (tok/s) | 59 | 88 |
| Latency (ms) | 691 | 716 |
| Vision | ||
| Tools | ||
| Reasoning | ||
| JSON mode |
Which should you pick?
DeepSeek: DeepSeek V4 Flash 0423 costs less per million input tokens — about 24.2× cheaper than OpenAI: GPT-5.6 Terra. OpenAI: GPT-5.6 Terra scores higher on measured quality. OpenAI: GPT-5.6 Terra takes a larger context window, which matters for long documents or big codebases. Prices here are live list prices; the same model can cost less on a different host, so check the model page for the cheapest offering.