Output tokens cost $0.02 per million on DeepSeek-OCR and $1.09 per million on DeepSeek-V3. Input tokens cost $0.02 per million on DeepSeek-OCR and $0.27 per million on DeepSeek-V3. Cached input tokens are billed at $0.02 per million on DeepSeek-OCR and $0.27 per million on DeepSeek-V3. Context lengths are 8,000 tokens on DeepSeek-OCR and 1,638,000 tokens on DeepSeek-V3. Time to first token (TTFT) measured on AIHubMix is 1.7s on DeepSeek-OCR and 1.2s on DeepSeek-V3. Measured output throughput is 105.2 tok/s on DeepSeek-OCR and 66.5 tok/s on DeepSeek-V3.
DeepSeek-OCRDeepSeek-V3DeepSeek-OCR is a vision-language model launched by DeepSeek AI, focusing on optical character recognition (OCR) and “contextual optical compression.” The model is designed to explore the limits of compressing contextual information from images, efficiently processing documents and converting them into structured text formats such as Markdown. The model requires an image as input.
It has been automatically upgraded to the latest released version, 250324. Automatically upgraded to the latest released version 250324.
Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.
Promotional prices show the discounted rate; see each model page for promotion windows.
Daily traffic served through AIHubMix — how demand for each model is trending.
Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.
Estimate your monthly bill for the same workload on each model.
Monthly = daily × 30. Discounted rates applied where a promotion is active.
DeepSeek-OCR: $0.02/M output tokens; DeepSeek-V3: $1.09/M. Use the cost calculator above to estimate your own workload.
DeepSeek-V3: 1.2s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.
DeepSeek-OCR accepts 8,000 and DeepSeek-V3 accepts 1,638,000 input tokens.
DeepSeek-OCR at 105.2 tok/s and DeepSeek-V3 at 66.5 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.
DeepSeek-OCR accepts text and image input; DeepSeek-V3 accepts text input and supports tool calling, function calling and structured outputs.
Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.
Related model match-ups readers also look at.
AIHubMix