D
DeepSeek V3.1OPEN
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates.
Prices update automatically — checked daily against provider list prices.
- $0.43 per 1M tokens, assuming 3 input for 1 output the convention used by Artificial Analysis
- cheaper than 40% of the 200 open-weight models tracked here
- 9 of 15 DeepSeek models tracked here, from cheapest
Benchmarks & performance
Independent benchmark scores for DeepSeek V3.1, measured by Artificial Analysis. Higher is better. Measurement mode: Reasoning.
About this model
DeepSeek V3.1 is an open-weight AI model by DeepSeek. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Frequently asked questions
What is DeepSeek V3.1?
DeepSeek V3.1 is an AI language model from DeepSeek. It is open-weight: you can download it and run it on your own hardware, for free.
Is DeepSeek V3.1 free?
The weights are free and open — you can self-host DeepSeek V3.1 and pay nothing per token. If you prefer a hosted API, list prices are $0.25 per million input tokens and $0.95 per million output tokens.
What is DeepSeek V3.1 good at?
Independent benchmarks from Artificial Analysis give it GPQA 77.9%, MMLU-Pro 85.1%, Humanity's Last Exam 14.3%, Long Context Reasoning 56.7%, LiveCodeBench 78.4%, AIME 2025 89.7%, IFBench 41.5%, τ²-Bench 37.4%, Terminal-Bench Hard 25%. It is particularly used for mathematical reasoning.
Can I self-host DeepSeek V3.1?
Yes. DeepSeek V3.1 has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.