The AI Agent Index

Documenting the technical and safety features of deployed agentic AI systems

DeepSeek-V3


Basic information

Website: https://web.archive.org/web/20241227082440/https://www.deepseek.com/

Short description: Deepseek-v3 is an open-source general purpose reasoning agent.

Intended uses: What does the developer say it’s for? General purpose uses including chat.

Date(s) deployed: Announced December 26, 2024 [source]


Developer

Website: https://web.archive.org/web/20241227082440/https://www.deepseek.com/

Legal name: DeepSeek [source]

Entity type: Corporation [source]

Country (location of developer or first author’s first affiliation): China [source]

Safety policies: What safety and/or responsibility policies are in place? Unknown


System components

Backend model: What model(s) are used to power the system? DeepSeek-V3 [source]

Publicly available model specification: Is there formal documentation on the system’s intended uses and how it is designed to behave in them? None

Reasoning, planning, and memory implementation: How does the system ‘think’? “We introduce an innovative methodology to distill reasoning capabilities from the long-Chain-of-Thought (CoT) model, specifically from one of the DeepSeek R1 series models, into standard LLMs, particularly DeepSeek-V3. Our pipeline elegantly incorporates the verification and reflection patterns of R1 into DeepSeek-V3 and notably improves its reasoning performance. Meanwhile, we also maintain a control over the output style and length of DeepSeek-V3.” [source]

Observation space: What is the system able to observe while ‘thinking’? Textual inputs from users. The DeepSpeek web chat app cannot search the web [source]

Action space/tools: What direct actions can the system take? Natural language.

User interface: How do users interact with the system? The DeepSpeek web chat app is a simple chat interface [source]. However, since the system is open-source, developers can interact through it with coding tools and build custom interfaces/scaffolding [source]

Development cost and compute: What is known about the development costs? Reported to take 6 million in training compute cost [source]


Guardrails and oversight

Accessibility of components:

  • Weights: Are model parameters available? Available [source]
  • Data: Is data available? Closed-source
  • Code: Is code available? Available [source]
  • Scaffolding: Is system scaffolding available? Available [source]
  • Documentation: Is documentation available? Available [source] in addition to a technical report [source]

Controls and guardrails: What notable methods are used to protect against harmful actions? None

Customer and usage restrictions: Are there know-your-customer measures or other restrictions on customers? None

Monitoring and shutdown procedures: Are there any notable methods or protocols that allow for the system to be shut down if it is observed to behave harmfully? None


Evaluation

Notable benchmark evaluations: Various reported [source] [source]

Bespoke testing: None

Safety: Have safety evaluations been conducted by the developers? What were the results? None

Publicly reported external red-teaming or comparable auditing:

  • Personnel: Who were the red-teamers/auditors? None
  • Scope, scale, access, and methods: What access did red-teamers/auditors have and what actions did they take? None
  • Findings: What did the red-teamers/auditors conclude? None

Ecosystem information

Interoperability with other systems: What tools or integrations are available? None

Usage statistics and patterns: Are there any notable observations about usage? Unknown


Additional notes