Intelligence index
Outperforms 47% of compared models
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.
- Input / output price
- $0.27/$0.41per 1M tokens
- Context
- 164K
- Released
- 2025-09-29
Pricing
Artificial Analysis
Explore rankingsDesign Arena
Elo scores and win rates across design and agent tasks.
Category performance
1,000-1,500 EloModel Arena
3D
1145Elo
Win rate 56.4%
UI Component
1161Elo
Win rate 53.3%
Code Categories
1170Elo
Win rate 54.2%
Game Development
1144Elo
Win rate 53%
Website
1177Elo
Win rate 54.3%
Data Visualization
1152Elo
Win rate 50.6%
SVG
1028Elo
Win rate 42%
Related Models
DeepSeek V4.1 Flash 1017
deepseekDeepSeek V4.1 Flash 1017 is the October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the deepseek-flash API model name. V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on output from a 552B-parameter backbone, an asymmetric split that keeps per-token compute low relative to the model's total size. Image understanding is native to the architecture, with visual and text embeddings trained jointly from the start of pre-training rather than added afterward as in the earlier experimental V4 Flash Vision Exp. It is suited for coding, terminal, and computer-use agents, along with long-horizon tasks that must run to completion across many steps and long-context analysis. Compressed KV caching cuts cache memory to roughly a quarter of the previous Flash generation, significantly reducing costs on agentic workloads. DeepSeek positions it as the cost-efficient tier of the V4.1 family and reports that it exceeds V4 Pro on performance, speed, and task completion time.
- Context
- 1.0M
- Max output
- 393.2K
- Released
- 2026-10-17
- Input
- $0.15/M
- Output
- $0.6/M
- Cache read
- $0.003/M
DeepSeek V4 Pro 0813
deepseekDeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro.
- Context
- 1.0M
- Max output
- 393.2K
- Released
- 2026-08-12
- Input
- $0.66/M
- Output
- $1.98/M
- Cache read
- $0.022/M
DeepSeek V4 Flash Vision Exp
deepseekDeepSeek V4 Flash Vision Exp is an experimental vision-enabled version of DeepSeek V4 Flash 0731 from DeepSeek, adding image understanding while matching the base model on text capabilities including agents, reasoning, and world knowledge. It is a sparse mixture-of-experts model with 13B active parameters out of 284B total. It is suited for document and chart understanding, visual question answering, and multimodal agent workflows that interleave text and images.
- Context
- 1.0M
- Max output
- 262.1K
- Released
- 2026-08-21
- Input
- $0.216/M
- Output
- $0.647/M
- Cache read
- $0.0069/M