DeepSeek V3.2 Exp
DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism designed to improve training and inference efficiency in long-context scenarios while maintaining output quality. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. Learn more in our docs The model was trained under conditions aligned with V3.1-Terminus to enable direct comparison. Benchmarking shows performance roughly on par with V3.1 across reasoning, coding, and agentic tool-use tasks, with minor tradeoffs and gains depending on the domain. This release focuses on validating architectural optimizations for extended context lengths rather than advancing raw task accuracy, making it primarily a research-oriented model for exploring efficient transformer designs.
deepseek/deepseek-v3.2-exp
- Context
- 164k tokens
- Completion cap
- 65,536
- Tools
- Yes
- JSON
- Yes
- Released
- 2025-09-29
Call it from Geek Hub
Same OpenAI SDK. Change the base URL and the model id.
curl https://api.geekhub.mx/v1/chat/completions \
-H "Authorization: Bearer $GEEKHUB_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek/deepseek-v3.2-exp",
"messages": [{"role": "user", "content": "Hola"}]
}'Get an API keyFrequently asked questions
- What is DeepSeek V3.2 Exp?
- DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism designed to improve training and inference efficiency in long-context scenarios while maintaining output quality. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. Learn more in our docs The model was trained under conditions aligned with V3.1-Terminus to enable direct comparison. Benchmarking shows performance roughly on par with V3.1 across reasoning, coding, and agentic tool-use tasks, with minor tradeoffs and gains depending on the domain. This release focuses on validating architectural optimizations for extended context lengths rather than advancing raw task accuracy, making it primarily a research-oriented model for exploring efficient transformer designs. DeepSeek V3.2 Exp runs on the Geek Hub API (OpenAI-compatible). Model id: deepseek/deepseek-v3.2-exp.
- Is DeepSeek V3.2 Exp free?
- No. Input is $0.2943 per 1M tokens and output is $0.4469 per 1M tokens on Geek Hub (markup included).
- What is the context length of DeepSeek V3.2 Exp?
- DeepSeek V3.2 Exp has a 164k tokens context window. It supports up to 65,536 completion tokens.
- Does DeepSeek V3.2 Exp support tool calling and structured outputs?
- DeepSeek V3.2 Exp accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
- When was DeepSeek V3.2 Exp released?
- DeepSeek V3.2 Exp was released on 2025-09-29.
More models from DeepSeek
DeepSeek V4 FlashV4 Flash open-weight, aggressive pricing with 1M context. Cache hit drops input to $0.0028. Replaces DeepSeek Chat V3.
DeepSeek V4 ProV4 Pro with reasoning + non-reasoning modes. Cache hit $0.003625. Replaces DeepSeek Reasoner R1.
DeepSeek V3DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tok
DeepSeek V3 0324DeepSeek V3, a 685B-parameter, mixture-of-experts model, is the latest iteration of the flagship chat model family from the DeepSeek team. It succeeds the [DeepSeek V3](/deepseek/d
DeepSeek V3.1DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 ba
DeepSeek V3.1 TerminusDeepSeek-V3.1 Terminus is an update to [DeepSeek V3.1](/deepseek/deepseek-chat-v3.1) that maintains the model's original capabilities while addressing issues reported by users, inc