AI Organization23 models

DeepSeek: AI models, benchmarks & research

Chinese AI company developing open and frontier language models, including the DeepSeek-V3 and DeepSeek-R1 model families.

Open-weight
#1factuality#3roleplay#5language#9math#9frontend development
Model
DeepSeek-V4-Flash-Max
DeepSeek-V4-Pro-Max
DeepSeek-V3.2
DeepSeek-V3.2 (Non-thinking)
DeepSeek-V3.2-Speciale
DeepSeek-V3.2 (Thinking)
DeepSeek-V3.2-Exp
DeepSeek-R1-0528
DeepSeek-V3 0324
DeepSeek-R1
DeepSeek R1 Distill Llama 70B
DeepSeek R1 Distill Llama 8B
DeepSeek R1 Distill Qwen 14B
DeepSeek R1 Distill Qwen 1.5B
DeepSeek R1 Distill Qwen 32B
DeepSeek R1 Distill Qwen 7B
DeepSeek R1 Zero
DeepSeek-V3.1
DeepSeek-V3
DeepSeek VL2Multimodal
DeepSeek VL2 SmallMultimodal
DeepSeek VL2 TinyMultimodal
DeepSeek-V2.5
At a glance

DeepSeekmodels, benchmarks & research

Key facts about DeepSeek's portfolio of 23 AI models — sourced from public benchmarks, provider APIs, and model releases.

Models
23 models across 0 families
Portfolio
23 language models
Licensing
23 open-weight, 0 proprietary
Benchmarks
Evaluated on 70 benchmarks
Cheapest model
DeepSeek-V4-Flash-Max at $0.10 per 1M input tokens
Active since
Shipping models since May 2024, latest release April 2026

FAQ

Common questions about DeepSeek.

What is DeepSeek?

DeepSeek is an AI research organization that develops large language models and other AI systems. They have released 23 models across 0 model families.

How many AI models has DeepSeek released?

DeepSeek has released 23 AI models tracked on LLM Stats. See the Models tab for the full catalog with technical specifications, benchmark scores, and pricing.

Does DeepSeek release open-weight models?

Yes. 23 of DeepSeek's 23 models are released under open-weight licenses. Check the Models tab for per-model license details.

How do DeepSeek's models perform on benchmarks?

DeepSeek's models have been evaluated on 70 benchmarks. Their top-performing model is DeepSeek-V4-Flash-Max on CodeForces with a score of 1.0. See the Benchmarks tab for the full breakdown.

What does it cost to use DeepSeek's models?

DeepSeek's most affordable model is DeepSeek-V4-Flash-Max with input pricing from $0.10 per 1M tokens via commercial API providers. 3 of their models have public pricing. See the Pricing tab for the full comparison.

When did DeepSeek start releasing AI models?

DeepSeek has been releasing AI models since May 2024. Their latest release was in April 2026. See the Timeline tab for the full release history.

How can I use DeepSeek's models?

Visit https://deepseek.com for official documentation and API access. Many of DeepSeek's models are also available through third-party API providers — check the Pricing tab to compare pricing across providers.