gpt-4o vs deepseek-v4-flash
OpenAI against DeepSeek, priced on the same four workloads with price book 2026-10-22.
On real workloads
| Workload | gpt-4o per call | deepseek-v4-flash per call | Monthly difference |
|---|---|---|---|
| Chat reply 1,500 input and 400 output tokens, 100,000 calls a month |
$0.00775 | $0.00093 | −$682.00 |
| RAG answer, cached context 2,000 fresh + 4,000 cached input, 500 output, 100,000 calls a month |
$0.01 | $0.00122 | −$1,377.60 |
| Agent step 4,000 fresh + 8,000 cached input, 800 output, 250,000 calls a month |
$0.03 | $0.00221 | −$6,448.00 |
| Batch summary 3,000 input and 300 output tokens via the Batch API, 1,000,000 calls a month |
$0.00525 | — | — |
Monthly difference is deepseek-v4-flash minus gpt-4o. Batch rows use each provider's Batch API rates where published. Price both on your own workload.
Rates
| Model | Input | Output | Cache read | Cache write | Context |
|---|---|---|---|---|---|
| openai/gpt-4o | $2.50 | $10.00 | $1.25 | — | 128K |
| deepseek/deepseek-v4-flash | $0.30 | $1.20 | $0.006 | $0.00 | 1M |
Price is one input. Before switching, run your own prompts through both and compare quality; then veqen diff prices the change at your production volume.