DeepSeek V4.1 Flash 1017 vs GPT-5.6 Luna
GPT-5.6 Luna is 1.2× cheaper for typical traffic ; DeepSeek V4.1 Flash 1017 scores higher on the Artificial Analysis intelligence index (39.5 vs 37.3).
Prices checked Oct 21, 2026
| DeepSeek V4.1 Flash 1017 | GPT-5.6 Luna | |
|---|---|---|
| Developer | DeepSeek | OpenAI |
| Input / 1M tokens | $0.30 | $0.20 |
| Output / 1M tokens | $1.20 | $1.20 |
| Cached input / 1M | $0.0060 | $0.02 |
| Context window | 1.05M | 1.05M |
| Max output | 393K | 128K |
| Intelligence index | 39.5 | 37.3 |
| Image input | Yes | Yes |
| Tool calling | Yes | Yes |
| Reasoning mode | Yes | Yes |
| Open weights | Yes | No |
Monthly cost for common workloads
| Workload | DeepSeek V4.1 Flash 1017 | GPT-5.6 Luna |
|---|---|---|
| Customer chatbot 1,000 conversations a day, 8 turns each | $145 | $134 |
| Document Q&A (RAG) 500 questions a day over retrieved documents | $33.35 | $25.38 |
| Coding / tool-using agent 200 tasks a day, long tool-heavy contexts | $57.42 | $52.20 |
| Content generation 30 long articles a day | $3.51 | $3.24 |
| Classification / extraction 100,000 short items a day | $324 | $234 |
Estimates include cached-input discounts where a model offers them. Try your own numbers in the cost calculator.
Which one to choose
- Lowest cost: GPT-5.6 Luna — 1.2× cheaper on a 3:1 input/output mix.
- Longer documents: GPT-5.6 Luna reads up to 1.05M tokens at once.
- Capabilities: DeepSeek V4.1 Flash 1017 — vision, tools, reasoning, JSON, open weights; GPT-5.6 Luna — vision, tools, reasoning, JSON.