# deepseek-ai/DeepSeek-R1

Repository: https://github.com/deepseek-ai/DeepSeek-R1
Canonical: https://ross.abutalabs.com/products/deepseek-r1
License: MIT
License Family: permissive
Last push: 2025-06-27T08:35:54+00:00

## Health v2 (maintenance only)
Score: 24/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 28, release rhythm 8, longevity 42
- inputs: {"age_days": 590, "days_push": 432, "days_rel": 432, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 92038, forks 11702 (observed 2026-08-28T04:12:23.405551+00:00)

## What it is
DeepSeek-R1 is a family of open-weight large language models trained with large-scale reinforcement learning for reasoning, including DeepSeek-R1-Zero, DeepSeek-R1, and six distilled dense models based on Llama and Qwen. The repository provides model weights, usage recommendations, and evaluation details for running the models locally or via API.

## Use cases
- run a reasoning llm locally
- download open-weight reasoning model weights
- solve math and coding problems with an open model
- fine-tune or build on distilled reasoning models
- compare open models against OpenAI o1
- self-host a chat model with strong reasoning

## When to choose
- you need an open-weight model with strong math, code, and reasoning performance
- you want to study or reproduce RL-based reasoning training
- you need smaller distilled models that run on modest hardware

## When to avoid
- you need a ready-made application or chat UI rather than model weights
- you lack the GPU resources to run large models locally
- you need multimodal capabilities beyond text reasoning

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, machine-learning, deep-learning, llm-training
- domain: large-language-models, artificial-intelligence, deep-learning, reinforcement-learning
- platform: python, cross-platform
- tags: reasoning-model, open-weights, reinforcement-learning, distilled-models, chain-of-thought, gpu, linux

## Member repositories
- deepseek-ai/DeepSeek-R1 (main) score 24

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:23.405551+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:12:52.798279+00:00, confidence not recorded.
  - readme: https://github.com/deepseek-ai/DeepSeek-R1 (fetched 2026-08-28T04:12:23.405551+00:00, sha ba6627faf0f9)
- Data as of 2026-08-30T08:39:29.467469+00:00.