DeepSeek
company
Verified
AI & ML interests
None defined yet.
Recent Activity
Papers
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
A Programming Paradigm for Spatiotemporal Composability
-
Chat with DeepSeek-VL2-small
🌍612Chat with images using AI for answers and visual tasks
-
deepseek-ai/deepseek-vl2-tiny
Image-Text-to-Text • 3B • Updated • 111k • 273 -
deepseek-ai/deepseek-vl2-small
Image-Text-to-Text • 16B • Updated • 5.25k • 195 -
deepseek-ai/deepseek-vl2
Image-Text-to-Text • 27B • Updated • 5.46k • 410
DeepSeek-Prover-Series
-
deepseek-ai/DeepSeek-Coder-V2-Instruct
Text Generation • 236B • Updated • 3.71k • 718 -
deepseek-ai/DeepSeek-Coder-V2-Base
Text Generation • 236B • Updated • 4.32k • 97 -
deepseek-ai/DeepSeek-Coder-V2-Lite-Base
Text Generation • 16B • Updated • 7.64k • 149 -
deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct
Text Generation • 16B • Updated • 91.8k • 712
DeepSeek-VL model series
DeepSeek LLM series
DeepSeek MoE series
-
deepseek-ai/DeepSeek-V3.2-Exp
Text Generation • 685B • Updated • 45.3k • • 1.03k -
deepseek-ai/DeepSeek-V3.2-Exp-Base
Text Generation • 685B • Updated • 975 • 68 -
deepseek-ai/DeepSeek-V3.2
Text Generation • 685B • Updated • 7.87M • • 1.6k -
deepseek-ai/DeepSeek-V3.2-Speciale
Text Generation • 685B • Updated • 3.78k • 746
-
deepseek-ai/DeepSeek-R1
Text Generation • 684B • Updated • 1.64M • • 14.4k -
deepseek-ai/DeepSeek-R1-Zero
Text Generation • 684B • Updated • 3.86k • 986 -
deepseek-ai/DeepSeek-R1-Distill-Llama-70B
Text Generation • 71B • Updated • 27.2k • • 824 -
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
Text Generation • 33B • Updated • 359k • • 1.67k
DeepSeek Math series
-
deepseek-ai/DeepSeek-Math-V2
Text Generation • 685B • Updated • 1.13k • 721 -
deepseek-ai/deepseek-math-7b-instruct
Text Generation • Updated • 2.75k • 168 -
deepseek-ai/deepseek-math-7b-rl
Text Generation • 7B • Updated • 1.9k • 110 -
deepseek-ai/deepseek-math-7b-base
Text Generation • Updated • 3.61k • 108
Janus is a novel autoregressive framework that unifies multimodal understanding and generation.
models for paper expert-specialized fine-tuning
-
deepseek-ai/ESFT-vanilla-lite
Text Generation • 16B • Updated • 1.14k • 34 -
deepseek-ai/ESFT-token-law-lite
Text Generation • 1B • Updated • 1.08k • 17 -
deepseek-ai/ESFT-token-summary-lite
Text Generation • 1B • Updated • 1.41k • 25 -
deepseek-ai/ESFT-token-code-lite
Text Generation • 2B • Updated • 1.1k • 18
DeepSeek Coder series
-
deepseek-ai/deepseek-coder-33b-instruct
Text Generation • 33B • Updated • 11k • 606 -
deepseek-ai/deepseek-coder-6.7b-instruct
Text Generation • 7B • Updated • 24.6k • 538 -
deepseek-ai/deepseek-coder-7b-instruct-v1.5
Text Generation • 7B • Updated • 161k • 193 -
deepseek-ai/deepseek-coder-1.3b-instruct
Text Generation • 1B • Updated • 33.1k • 211
-
deepseek-ai/DeepSeek-V3.2-Exp
Text Generation • 685B • Updated • 45.3k • • 1.03k -
deepseek-ai/DeepSeek-V3.2-Exp-Base
Text Generation • 685B • Updated • 975 • 68 -
deepseek-ai/DeepSeek-V3.2
Text Generation • 685B • Updated • 7.87M • • 1.6k -
deepseek-ai/DeepSeek-V3.2-Speciale
Text Generation • 685B • Updated • 3.78k • 746
-
deepseek-ai/DeepSeek-R1
Text Generation • 684B • Updated • 1.64M • • 14.4k -
deepseek-ai/DeepSeek-R1-Zero
Text Generation • 684B • Updated • 3.86k • 986 -
deepseek-ai/DeepSeek-R1-Distill-Llama-70B
Text Generation • 71B • Updated • 27.2k • • 824 -
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
Text Generation • 33B • Updated • 359k • • 1.67k
DeepSeek Math series
-
deepseek-ai/DeepSeek-Math-V2
Text Generation • 685B • Updated • 1.13k • 721 -
deepseek-ai/deepseek-math-7b-instruct
Text Generation • Updated • 2.75k • 168 -
deepseek-ai/deepseek-math-7b-rl
Text Generation • 7B • Updated • 1.9k • 110 -
deepseek-ai/deepseek-math-7b-base
Text Generation • Updated • 3.61k • 108
-
Chat with DeepSeek-VL2-small
🌍612Chat with images using AI for answers and visual tasks
-
deepseek-ai/deepseek-vl2-tiny
Image-Text-to-Text • 3B • Updated • 111k • 273 -
deepseek-ai/deepseek-vl2-small
Image-Text-to-Text • 16B • Updated • 5.25k • 195 -
deepseek-ai/deepseek-vl2
Image-Text-to-Text • 27B • Updated • 5.46k • 410
Janus is a novel autoregressive framework that unifies multimodal understanding and generation.
DeepSeek-Prover-Series
-
deepseek-ai/DeepSeek-Coder-V2-Instruct
Text Generation • 236B • Updated • 3.71k • 718 -
deepseek-ai/DeepSeek-Coder-V2-Base
Text Generation • 236B • Updated • 4.32k • 97 -
deepseek-ai/DeepSeek-Coder-V2-Lite-Base
Text Generation • 16B • Updated • 7.64k • 149 -
deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct
Text Generation • 16B • Updated • 91.8k • 712
models for paper expert-specialized fine-tuning
-
deepseek-ai/ESFT-vanilla-lite
Text Generation • 16B • Updated • 1.14k • 34 -
deepseek-ai/ESFT-token-law-lite
Text Generation • 1B • Updated • 1.08k • 17 -
deepseek-ai/ESFT-token-summary-lite
Text Generation • 1B • Updated • 1.41k • 25 -
deepseek-ai/ESFT-token-code-lite
Text Generation • 2B • Updated • 1.1k • 18
DeepSeek-VL model series
DeepSeek Coder series
-
deepseek-ai/deepseek-coder-33b-instruct
Text Generation • 33B • Updated • 11k • 606 -
deepseek-ai/deepseek-coder-6.7b-instruct
Text Generation • 7B • Updated • 24.6k • 538 -
deepseek-ai/deepseek-coder-7b-instruct-v1.5
Text Generation • 7B • Updated • 161k • 193 -
deepseek-ai/deepseek-coder-1.3b-instruct
Text Generation • 1B • Updated • 33.1k • 211
DeepSeek LLM series
DeepSeek MoE series