DeepSeek

About DeepSeek

深度求索(DeepSeek),探索未至之境。专注于研究世界领先的通用人工智能底层模型与技术,开源 DeepSeek-V4、DeepSeek-R1 等前沿大模型。和 DeepSeek AI 对话,轻松接入 API。

Website ↗GitHub ↗

Source ↗ · 2026-11-12T02:17:03.411506080Z

Transparency & Governance

  • Foundation Model Transparency Index 2025 · 32 / 100 · Company disclosure assessment Source assessment ↗

101 indexed models

101 matching models

DeepSeek: DeepSeek-R1-0120

siliconFlow/deepseek-ai/DeepSeek-R1-0120

DeepSeek-R1 is a reasoning model powered by reinforcement learning (RL) that addresses the issues of repetition and readability. Prior to RL, DeepSeek-R1 incorporated cold-start data to further optimize its reasoning performance. It achieves performance comparable to OpenAI-o1 ac

Input modalities
text
Output modalities
text
Capabilities
structuredOutputs, toolCalling

DeepSeek: DeepSeek-R1-Distill-Llama-70B

siliconFlow/deepseek-ai/DeepSeek-R1-Distill-Llama-70B

DeepSeek-R1-Distill-Llama-70B is a distilled model based on Llama-3.3-70B-Instruct. As part of the DeepSeek-R1 series, it was fine-tuned using samples generated by DeepSeek-R1 and demonstrates excellent performance across mathematics, programming, and reasoning tasks. The model a

Input modalities
text
Output modalities
text

Benchmarks

  • LiveBench: 42.294%
  • LLM Stats · GitHub archive: 0.652
  • Open LLM Leaderboard v2: 27.809%

DeepSeek: DeepSeek-R1-Distill-Llama-8B

siliconFlow/deepseek-ai/DeepSeek-R1-Distill-Llama-8B

DeepSeek-R1-Distill-Llama-8B is a distilled model based on Llama-3.1-8B. The model was fine-tuned using samples generated by DeepSeek-R1 and demonstrates strong reasoning capabilities. It achieved notable results across various benchmarks, including 89.1% accuracy on MATH-500, 50

Input modalities
text
Output modalities
text

Benchmarks

  • LLM Stats · GitHub archive: 0.49
  • Open LLM Leaderboard v2: 13.06%

DeepSeek: DeepSeek-R1-Distill-Qwen-14B

siliconFlow/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B

DeepSeek-R1-Distill-Qwen-14B is a distilled model based on Qwen2.5-14B. The model was fine-tuned using 800k curated samples generated by DeepSeek-R1 and demonstrates strong reasoning capabilities. It achieved impressive results across various benchmarks, including 93.9% accuracy

Maximum output
131,072
Input modalities
text
Output modalities
text
Capabilities
structuredOutputs, toolCalling

Benchmarks

  • LLM Stats · GitHub archive: 0.591
  • Open LLM Leaderboard v2: 38.221%

DeepSeek: DeepSeek-R1-Distill-Qwen-7B

siliconFlow/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B

DeepSeek-R1-Distill-Qwen-7B is a distilled model based on Qwen2.5-Math-7B. The model was fine-tuned using 800k curated samples generated by DeepSeek-R1 and demonstrates strong reasoning capabilities. It achieved impressive results across various benchmarks, including 92.8% accura

Maximum output
16,384
Input modalities
text
Output modalities
text
Capabilities
structuredOutputs, toolCalling

Benchmarks

  • LLM Stats · GitHub archive: 0.491
  • Open LLM Leaderboard v2: 14.995%

DeepSeek: DeepSeek V3

deepseek/deepseek-chat-v3

DeepSeek-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 tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals

Context window
163,840
Input modalities
text
Output modalities
text
Capabilities
structuredOutputs, toolCalling

Benchmarks

  • HELM Capabilities: 83.052%
  • LiveBench: 54.668%
  • LLM Stats · GitHub archive: 0.885
  • OpenRouter Benchmarks: 40.38%

Inference prices

  • Input tokens (per 1M) : $0.25 / input_tokens_per_million · siliconFlow ↗
  • Output tokens (per 1M) : $1.00 / output_tokens_per_million · siliconFlow ↗

DeepSeek: DeepSeek V4.1 Flash

deepseek/deepseek-v4.1-flash-20260910

DeepSeek 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 c

Context window
1,048,576
Input modalities
text, image
Output modalities
text
Capabilities
multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput

Benchmarks

  • Artificial Analysis: 24.7
  • OpenRouter Benchmarks: 90.78%

Inference prices

  • Published price : $0.0042 / cache_read_tokens_per_million · openRouter ↗
  • Published price : $0.14 / input_tokens_per_million · openRouter ↗
  • Published price : $0.42 / output_tokens_per_million · openRouter ↗

DeepSeek: DeepSeek-VL2

siliconFlow/deepseek-ai/deepseek-vl2

DeepSeek-VL2 is a mixed-expert (MoE) vision-language model developed based on DeepSeekMoE-27B, employing a sparse-activated MoE architecture to achieve superior performance with only 4.5B active parameters. The model excels in various tasks including visual question answering, op

Maximum output
4,096
Input modalities
text, image
Output modalities
text
Capabilities
multimodal, structuredOutputs, visionInput

Benchmarks

  • MMBench: 87.1%

DeepSeek-OCR

vertexAI/publishers/deepseek-ai/models/deepseek-ocr-maas@001

Input modalities
image, text
Output modalities
text
Capabilities
multimodal, ocr, visionInput

deepseek-r1

cheaperInference/deepseek-r1

Context window
160,000
Maximum output
8,192
Input modalities
text
Output modalities
text
Capabilities
reasoning, toolCalling

Inference prices

DeepSeek-R1

awsBedrock/deepseek.r1-v1:0

Input modalities
text
Output modalities
text

deepseek-r1-0528

cheaperInference/deepseek-r1-0528

Context window
131,072
Maximum output
32,768
Input modalities
text
Output modalities
text
Capabilities
reasoning, structuredOutputs, toolCalling

Inference prices

DeepSeek R1 (0528)

vertexAI/publishers/deepseek-ai/models/deepseek-r1-0528-maas@001

Input modalities
text
Output modalities
text
Capabilities
batch, structuredOutputs, toolCalling

DeepSeek: R1 0528

deepseek/deepseek-r1-0528

May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass. Fully open-source model.

Context window
163,840
Input modalities
text
Output modalities
text
Capabilities
promptCaching, reasoning, structuredOutputs, toolCalling

Benchmarks

  • Artificial Analysis: 13.1
  • HELM Capabilities: 66.592%
  • OpenCompass LLM: 83.5%
  • OpenRouter Benchmarks: 79.586%

Inference prices

  • Published price : $0.50 / input_tokens_per_million · deepInfra ↗
  • Published price : $2.15 / output_tokens_per_million · deepInfra ↗

deepseek-r1-community

cheaperInference/deepseek-r1-community

Context window
64,000
Maximum output
8,000
Input modalities
text
Output modalities
text
Capabilities
reasoning, toolCalling

Inference prices

deepseek-r1-q4_k_m-gguf

vertexAI/publishers/deepseek-ai/models/deepseek-r1@deepseek-r1-q4_k_m-gguf

Input modalities
text

deepseek-r1-turbo

cheaperInference/deepseek-r1-turbo

Context window
64,000
Maximum output
16,000
Input modalities
text
Output modalities
text
Capabilities
reasoning, structuredOutputs, toolCalling

Inference prices

deepseek-v3

cheaperInference/deepseek-v3

Context window
64,000
Maximum output
16,000
Input modalities
text
Output modalities
text
Capabilities
toolCalling

Inference prices

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace