# winfunc/deepreasoning

A high-performance LLM inference API and Chat UI that integrates DeepSeek R1's CoT reasoning traces with Anthropic Claude models.

Repository: https://github.com/winfunc/deepreasoning
Canonical: https://ross.abutalabs.com/products/deepreasoning
Language: Rust
License: MIT
License Family: permissive
Topics: ai, anthropic, anthropic-claude, api, chain-of-thought, claude, deepseek, deepseek-r1, llm, rust
Last push: 2025-10-07T05:28:22+00:00

## Health v2 (maintenance only)
Score: 31/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 45, release rhythm 8, longevity 41
- inputs: {"age_days": 584, "days_push": 330, "days_rel": 584, "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 5363, forks 437 (observed 2026-08-28T04:09:16.068124+00:00)

## What it is
DeepReasoning is a high-performance Rust-based LLM inference API and chat UI that combines DeepSeek R1's chain-of-thought reasoning traces with Anthropic Claude's creative and code generation capabilities in a single unified stream. It supports self-hosting and BYOK (bring your own key) usage, keeping API keys and data under user control.

## Use cases
- combine deepseek r1 reasoning with claude responses in one api
- self-host an llm chat interface with my own api keys
- get chain-of-thought reasoning traces followed by a claude answer
- stream reasoning and final answers from two llm providers together
- run a private chatgpt-style ui without sending data to third parties
- improve claude's math and logic answers with r1-style reasoning

## When to choose
- you want R1's reasoning quality combined with Claude's coding and conversational strengths
- you need a self-hosted, privacy-focused inference proxy with local key management
- you want a single streaming API that merges two model providers' outputs
- you prefer a high-performance Rust backend with a ready-made chat UI

## When to avoid
- you need fully local/offline inference since it depends on external DeepSeek and Anthropic APIs
- you want a single-provider solution without multi-model orchestration
- you require official vendor support or affiliation with Anthropic or DeepSeek
- you need fine-tuning or model training capabilities

## Facets
- artifact type: service
- maturity: active
- function: llm-inference, http-server, api-framework, chatbot, rag
- domain: large-language-models, artificial-intelligence, developer-tools
- platform: rust, self-hosted, cross-platform
- tags: deepseek-r1, anthropic-claude, chain-of-thought, reasoning, byok, chat-ui, streaming, ai-agents, web-server

## Member repositories
- winfunc/deepreasoning (main) score 31

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:16.068124+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 2026-08-29T17:58:44.768132+00:00, confidence not recorded.
  - readme: https://github.com/winfunc/deepreasoning (fetched 2026-08-28T04:09:16.068124+00:00, sha 26be9eb4fb27)
- Data as of 2026-08-30T08:39:29.467469+00:00.