DeepSeek V4 Flash is a preview Mixture-of-Experts language model in the DeepSeek V4 series. It has 284B parameters with 13B activated and supports a context length of one million tokens. The model uses a hybrid attention architecture (Compressed Sparse Attention and Heavily Compressed Attention), Manifold-Constrained Hyper-Connections, and the Muon optimizer. It was pre-trained on over 32T tokens and post-trained with a two-stage paradigm. Its repository is deepseek-ai/DeepSeek-V4-Flash, it is licensed under MIT, and its pipeline is text-generation.
Model size
290.94B
Format
unknown weights
Context
1,048,576
Freshness 51 out of 100. Repository age: 176 days old, Apr 22, 2026. Measures age, not quality.Freshness51/100176 days old
Popularity 81 out of 100. 805,960 downloads in 30 days; 2,324 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.Popularity81/100806K downloads / 30d
Arena Open Source Text: rank 25 of 225. Ranked model: deepseek-v4-flash. Export dated Oct 13, 2026. Source checked Oct 15, 2026. Human preference ranking, not hardware suitability.#25Arena · Open SourceText
DeepSeek V4 Flash 0731 is a text-generation model released under the MIT license on the deepseek-ai repository. It is described as the official release of DeepSeek-V4-Flash, superseding a preview version, with enhanced agentic capabilities. It shares the same model structure as DeepSeek-V4-Flash-DSpark and includes a speculative decoding module. The release does not include a Jinja-format chat template; instead, an encoding folder provides Python scripts and test cases for converting OpenAI-compatible messages to model inputs and parsing outputs. The reasoning_effort parameter supports three levels. Benchmark tables compare its reported scores with other models.
Model size
304.18B
Format
unknown weights
Context
1,048,576
Freshness 75 out of 100. Repository age: 76 days old, Jul 31, 2026. Measures age, not quality.Freshness75/10076 days old
Popularity 90 out of 100. 4,416,319 downloads in 30 days; 4,083 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.Popularity90/1004.4M downloads / 30d
DeepSeek V4 Flash Vision Exp from deepseek-ai. Source-based hardware guidance from its published configuration.
Model size
304.65B
Format
unknown weights
Context
1,048,576
Freshness 84 out of 100. Repository age: 45 days old, Aug 31, 2026. Measures age, not quality.Freshness84/10045 days old
Popularity 77 out of 100. 508,455 downloads in 30 days; 951 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.Popularity77/100508.5K downloads / 30d
DeepSeek V4 Pro is a preview Mixture-of-Experts text-generation model from the DeepSeek-V4 series, released under the MIT license with Safetensors and Transformers compatibility. It has 1.6T total parameters with 49B activated and supports a context length of one million tokens. The model card describes a hybrid attention architecture combining Compressed Sparse Attention and Heavily Compressed Attention, Manifold-Constrained Hyper-Connections, and Muon optimizer training. It was pre-trained on more than 32T tokens and post-trained with a two-stage paradigm involving domain-specific SFT and RL with GRPO followed by on-policy distillation. A DeepSeek-V4-Pro-Max mode is described as offering maximum reasoning effort.
Model size
1598.84B
Format
unknown weights
Context
1,048,576
Freshness 51 out of 100. Repository age: 176 days old, Apr 22, 2026. Measures age, not quality.Freshness51/100176 days old
Popularity 79 out of 100. 383,720 downloads in 30 days; 5,673 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.Popularity79/100383.7K downloads / 30d
Arena Open Source Text: rank 11 of 225. Ranked model: deepseek-v4-pro. Export dated Oct 13, 2026. Source checked Oct 15, 2026. Human preference ranking, not hardware suitability.#11Arena · Open SourceText+1
DeepSeek V4 Pro 0813 is an official text-generation model from DeepSeek AI, released under the MIT license. It builds on the DeepSeek-V4-Pro (Preview) model structure and adds a DSpark speculative decoding module. The model is intended to offer enhanced agentic capabilities and performance improvements, particularly in production environments.
Model size
1650.50B
Format
unknown weights
Context
1,048,576
Freshness 78 out of 100. Repository age: 63 days old, Aug 13, 2026. Measures age, not quality.Freshness78/10063 days old
Popularity 71 out of 100. 141,935 downloads in 30 days; 897 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.Popularity71/100141.9K downloads / 30d