# GPT vs DeepSeek: API prices compared, dated and sourced

Canonical page: https://bertrande.com/en/vs/gpt-vs-deepseek/
License: CC-BY-4.0, reuse with attribution and a link.

No value below was produced, estimated or completed by a language model. Every figure was read from a source we name, on the date we give.

The cheapest GPT model we track, gpt-5-nano, costs $0.400 per million output tokens and $0.050 per million input tokens. The cheapest DeepSeek model, deepseek-flash, costs $1.20 per million output tokens and $0.300 per million input tokens. At the entry level, GPT is 3.0 times cheaper than DeepSeek on output tokens. Prices read on 11 September 2026 and 18 September 2026.

## GPT models

| Model | Input $/1M | Output $/1M | Read on |
|---|---|---|---|
| gpt-5-nano | $0.050 | $0.400 | 2026-09-18 |
| gpt-4.1-nano | $0.100 | $0.400 | 2026-09-18 |
| gpt-4o-mini | $0.150 | $0.600 | 2026-09-18 |
| gpt-5.6-luna | $0.200 | $1.20 | 2026-09-18 |
| gpt-5.4-nano | $0.200 | $1.25 | 2026-09-18 |
| gpt-3.5-turbo | $0.500 | $1.50 | 2026-09-18 |
| gpt-3.5-turbo-0125 | $0.500 | $1.50 | 2026-09-18 |
| gpt-4.1-mini | $0.400 | $1.60 | 2026-09-18 |
| gpt-5-mini | $0.250 | $2.00 | 2026-09-18 |
| gpt-3.5-turbo-1106 | $1.00 | $2.00 | 2026-09-18 |
| gpt-3.5-turbo-instruct | $1.50 | $2.00 | 2026-09-18 |
| o3-mini | $1.10 | $4.40 | 2026-09-18 |
| o4-mini | $1.10 | $4.40 | 2026-09-18 |
| gpt-5.4-mini | $0.750 | $4.50 | 2026-09-18 |
| gpt-4.1 | $2.00 | $8.00 | 2026-09-18 |
| o3 | $2.00 | $8.00 | 2026-09-18 |
| gpt-5 | $1.25 | $10.00 | 2026-09-18 |
| gpt-5.1 | $1.25 | $10.00 | 2026-09-18 |
| gpt-4o | $2.50 | $10.00 | 2026-09-18 |
| gpt-5.6-terra | $2.00 | $12.00 | 2026-09-18 |
| gpt-5.2 | $1.75 | $14.00 | 2026-09-18 |
| gpt-5.4 (<272K context length) | $2.50 | $15.00 | 2026-09-18 |
| gpt-4o-2024-05-13 | $5.00 | $15.00 | 2026-09-18 |
| gpt-5.6-sol | $4.00 | $20.00 | 2026-09-18 |
| gpt-5.5 (<272K context length) | $5.00 | $30.00 | 2026-09-18 |
| gpt-4-turbo-2024-04-09 | $10.00 | $30.00 | 2026-09-18 |
| gpt-6-astra | $10.00 | $50.00 | 2026-09-18 |
| o1 | $15.00 | $60.00 | 2026-09-18 |
| gpt-4-0613 | $30.00 | $60.00 | 2026-09-18 |
| o3-pro | $20.00 | $80.00 | 2026-09-18 |
| gpt-5-pro | $15.00 | $120 | 2026-09-18 |
| gpt-5.2-pro | $21.00 | $168 | 2026-09-18 |
| gpt-5.4-pro (<272K context length) | $30.00 | $180 | 2026-09-18 |
| gpt-5.5-pro (<272K context length) | $30.00 | $180 | 2026-09-18 |
| o1-pro | $150 | $600 | 2026-09-18 |

## DeepSeek models

| Model | Input $/1M | Output $/1M | Read on |
|---|---|---|---|
| deepseek-flash | $0.300 | $1.20 | 2026-09-11 |
| deepseek-v4-pro | $1.32 | $3.96 | 2026-09-11 |


Method: https://bertrande.com/en/methodology/
Report an error: https://bertrande.com/en/corrections/