Create an API key

DeepSeek models, priced in toman

18 DeepSeek models are live on uttapen. Prices are in toman per 1 million tokens, and you are charged for the tokens you actually use.

DeepSeek on uttapen

DeepSeek currently has 18 models in the uttapen catalogue. All of them bill against a toman wallet, for the tokens a request actually used.

The spread inside DeepSeek alone is wide: DeepSeek V4 Flash 0423 runs about 123 toman for a 1,000-word request, while DeepSeek Pro Latest costs 3,642 toman for the same one. Picking the right model inside a brand matters as much as picking the brand.

Measured against the whole catalogue, the median DeepSeek output price undercuts about 70% of the other paid models.

On capabilities: 5 take image input, 18 support tool calling, 16 do multi-step reasoning. The longest context belongs to DeepSeek V4 Flash 0423, at 1,048,576 tokens.

Every DeepSeek model answers on the same sk-up- key and the same base_url: switching between them means changing the `model` string and nothing else.

Which DeepSeek model should I start with?

Every DeepSeek model, priced in toman

ModelCapabilitiesContextInput / 1MOutput / 1M
DeepSeek V3deepseek/deepseek-chatTool callingJSON output163,84089,117356,156
DeepSeek V3 0324deepseek/deepseek-chat-v3-0324Tool callingJSON output163,840100,404394,690
DeepSeek V3.1deepseek/deepseek-chat-v3.1Tool callingReasoningJSON output163,84086,555328,908
R1 0528deepseek/deepseek-r1-0528Tool callingReasoningJSON output163,840173,110744,372
DeepSeek V3.1 Terminusdeepseek/deepseek-v3.1-terminusTool callingReasoningJSON output163,84093,479346,219
DeepSeek V3.2deepseek/deepseek-v3.2Tool callingReasoningJSON output163,84090,017131,563
DeepSeek V3.2 Expdeepseek/deepseek-v3.2-expTool callingReasoningJSON output163,84093,479141,950
DeepSeek V4 Flash 0423deepseek/deepseek-v4-flashTool callingReasoningJSON output1,048,57631,50663,012
DeepSeek V4 Flash 0731deepseek/deepseek-v4-flash-0731Tool callingReasoningJSON output1,048,57620,427103,866
DeepSeek V4 Flash Vision Expdeepseek/deepseek-v4-flash-vision-expImage inputTool callingReasoningJSON output1,048,57674,645223,935
DeepSeek V4.1 Flash 1017deepseek/deepseek-v4.1-flash-1017Image inputTool callingReasoningJSON output1,048,5762,077173,110
DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flashImage inputTool callingReasoningJSON output1,048,57615,580173,110
DeepSeek V4 Pro 0813deepseek/deepseek-v4-pro-0813Tool callingReasoningJSON output1,048,57635,6611,422,962
DeepSeek V4 Pro 0423deepseek/deepseek-v4-proTool callingReasoningJSON output1,048,576553,9511,107,902
DeepSeek ProLatest~deepseek/deepseek-pro-latestTool callingReasoningJSON output1,048,57632,0952,769,755
DeepSeek FlashLatest~deepseek/deepseek-flash-latestImage inputTool callingReasoningJSON output1,048,5762,077173,110
DeepSeek V4.1 FlashBatchdeepseek/deepseek-v4.1-flash:batchImage inputTool callingReasoningJSON output1,048,57638,777116,330
DeepSeek V4 FlashLatest~deepseek/deepseek-v4-flash-latestTool callingReasoningJSON output1,048,57620,427103,866

Common questions about the DeepSeek API

How do I reach the DeepSeek API from Iran?
Sign up with your mobile number, top the wallet up in toman and create an sk-up- key. Then, in the official OpenAI SDK, point base_url at https://api.uttapen.ir/v1 and set model to the DeepSeek model id. No international card, no VPN.
What do DeepSeek models cost in toman?
It depends on the model: the cheapest, DeepSeek V4 Flash 0423, is about 123 toman for a 1,000-word request, and the dearest, DeepSeek Pro Latest, about 3,642 toman. Each model page carries its own exact price, and you are billed for what the request actually used.
How many DeepSeek models are on uttapen?
18. 18 support tool calling and 5 accept image input. The catalogue re-syncs hourly, so a newly released model shows up here on its own.

Full model catalogue·All providers