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-v3.1

cheaperInference/deepseek-v3.1

Context window
128,000
Maximum output
128,000
Input modalities
text
Output modalities
text
Capabilities
promptCaching, reasoning, structuredOutputs, toolCalling

Benchmarks

  • Artificial Analysis: 13.7
  • OpenRouter Benchmarks: 72.941%

Inference prices

DeepSeek-V3.1

vertexAI/publishers/deepseek-ai/models/deepseek-v3.1-maas@001

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

deepseek-v3.2

cheaperInference/deepseek-v3.2

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

Inference prices

DeepSeek V3.2

awsBedrock/deepseek.v3.2

Input modalities
text
Output modalities
text

DeepSeek-V3.2

vertexAI/publishers/deepseek-ai/models/deepseek-v3.2-maas@001

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

DeepSeek-V3.2 · azureml-deepseek v1

azureFoundry/azureml-deepseek/DeepSeek-V3.2@1

DeepSeek-V3.2, a model that harmonizes high computational efficiency with superior reasoning and agent performance

Context window
128,000
Maximum input
128,000
Maximum output
128,000
Input modalities
text
Output modalities
text
Capabilities
toolCalling

DeepSeek-V3.2-Speciale · azureml-deepseek v1

azureFoundry/azureml-deepseek/DeepSeek-V3.2-Speciale@1

DeepSeek-V3.2 Speciale, a model that harmonizes high computational efficiency with superior reasoning and agent performance

Context window
128,000
Maximum input
128,000
Maximum output
128,000
Input modalities
text
Output modalities
text
Capabilities
toolCalling

deepseek-v3.2-thinking

cheaperInference/deepseek-v3.2-thinking

Context window
128,000
Maximum output
8,000
Input modalities
text
Output modalities
text
Capabilities
promptCaching, reasoning, toolCalling

Inference prices

DeepSeek V3.2 Thinking

vercelAIGatewayCatalog/deepseek/deepseek-v3.2-thinking

DeepSeek‑V3.2 from DeepSeek harmonizes high computational efficiency with superior reasoning and agent performance. It builds on three main techniques: DeepSeek Sparse Attention for long‑context efficiency, a scalable reinforcement learning framework, and a large‑scale agentic ta

Context window
128,000
Maximum output
8,000
Input modalities
text
Output modalities
text

deepseek-v4.1-flash

cheaperInference/deepseek-v4.1-flash

Context window
1,048,576
Maximum output
393,216
Input modalities
text, image
Output modalities
text
Capabilities
multimodal, reasoning, structuredOutputs, toolCalling, visionInput

deepseek-v4.1-flash-1017

cheaperInference/deepseek-v4.1-flash-1017

Context window
1,048,576
Maximum output
393,216
Input modalities
text, image
Output modalities
text
Capabilities
multimodal, reasoning, structuredOutputs, toolCalling, visionInput

DeepSeek V4.1 Flash 1017

mistral/deepseek-v4-1-flash-1017

DeepSeek V4.1 Flash 1017 is a third-party open weight model from DeepSeek, hosted by Mistral, with support for text and image inputs. Ideal for long-context coding and agentic workflows. The model is served without Mistral modifications.

Context window
1,000,000
Maximum output
128,000
Input modalities
text, image
Output modalities
text
Capabilities
batch, multimodal, promptCaching, reasoning, streaming, structuredOutputs, toolCalling, visionInput

Inference prices

  • Cached input / 1M tokens : $0.01 / cache_read_tokens_per_million · mistral ↗
  • input / 1M tokens : $0.30 / input_tokens_per_million · mistral ↗
  • output / 1M tokens : $1.20 / output_tokens_per_million · mistral ↗

DeepSeek-V4.1-Flash-1017 · azureml-deepseek v2026-10-17

azureFoundry/azureml-deepseek/DeepSeek-V4.1-Flash-1017@2026-10-17

DeepSeek-V4.1-Flash-1017, a multimodal Mixture-of-Experts (MoE) model with 552B backbone parameters and support for contexts of up to one million tokens.

Context window
1,000,000
Maximum output
128,000
Input modalities
text
Output modalities
text
Capabilities
reasoning, toolCalling

deepseek-v4.1-flash-1017-fast

cheaperInference/deepseek-v4.1-flash-1017-fast

Context window
1,048,576
Maximum output
262,144
Input modalities
text, image
Output modalities
text
Capabilities
multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput

Inference prices

DeepSeek V4.1 Flash 1017 Fast

vercelAIGatewayCatalog/deepseek/deepseek-v4.1-flash-1017-fast

Higher-throughput variant of DeepSeek V4.1 Flash 1017 served by Baseten for multimodal reasoning, tool calling, and structured output, with automatic prefix caching.

Context window
1,048,576
Maximum output
262,144
Input modalities
text, image
Output modalities
text

deepseek-v4.1-flash-1017-ultrafast

cheaperInference/deepseek-v4.1-flash-1017-ultrafast

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

Inference prices

DeepSeek-V4.1-Flash · azureml-deepseek v2026-09-10

azureFoundry/azureml-deepseek/DeepSeek-V4.1-Flash@2026-09-10

DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts (MoE) model with 552B backbone parameters and support for contexts of up to one million tokens.

Context window
1,000,000
Maximum output
128,000
Input modalities
text
Output modalities
text
Capabilities
reasoning, toolCalling

deepseek-v4.1-flash-fast

cheaperInference/deepseek-v4.1-flash-fast

Context window
1,048,576
Maximum output
32,768
Input modalities
text, image
Output modalities
text
Capabilities
multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput

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