LLM & Foundation Models

Bionic GPT

bionic-gpt/bionic-gpt

Bionic is a Rust-based, on-premise alternative to ChatGPT designed for private generative AI deployments. It provides a familiar chat interface, local or remote model access, team controls, retrieval-augmented assistants, data integrations, observability, and Kubernetes-oriented scaling.

Open official repository ↗ Project homepage ↗
★ 2.4KStars
⑂ 239Forks
5Open issues
RustLanguage
NOASSERTIONLicense
Q@project.QualityScoreEditorial score

Overview

The project can run on a laptop for local evaluation or on Kubernetes for larger deployments. Its documented architecture combines an Nginx router, OAuth2 proxy, Rust web server, RAG and document-processing engines, PostgreSQL with PgVector, S3-compatible object storage, an external identity provider, and one or more LLM providers. The repository has 2,346 stars, 239 forks, 5 open issues, and is not archived.

Key features

  • ChatGPT-inspired interface with chat history and customizable themes
  • Locally hosted or data-center deployment
  • AI assistants backed by agentic retrieval-augmented generation pipelines
  • Document support including HTML, PDF, CSV, PNG, and PPTX
  • UI-based embedding and chunking configuration
  • Team management, SSO integration, RBAC, and team-level data isolation
  • Local and remote LLM support with multiple-model switching and parallel conversations
  • OpenAI-compatible APIs for assistants with API-key and throttling controls
  • Airbyte-based batch and real-time data integrations
  • Manual uploads, datasets, and OCR processing
  • Prometheus-compatible observability and Grafana dashboard support
  • Token limits enforced through a model reverse proxy
  • Batch and streaming guardrails
  • Kubernetes-oriented deployment and scaling

Requirements, installation and quick start

For local evaluation, the README directs users to the Docker Compose installation guide at https://bionic-gpt.com/docs/running-locally/docker-compose/. For production, it directs users to https://bionic-gpt.com/docs/. Set APP_BASE_URL to the public URL of the Bionic server for OAuth2 callback construction. Exact installation commands are not stated in the supplied repository record.

Usage

After deployment, connect the service to an identity provider and one or more supported LLM providers, then access the web interface to chat, create assistants, upload or integrate datasets, configure embeddings and chunking, share assistants with team members, and switch among available models. Assistants may also be exposed as OpenAI-compatible APIs. A complete quick-start command sequence is not stated in the repository metadata.

Model compatibility and use cases

The architecture explicitly names Ollama, OpenAI, and Anthropic as example LLM providers. The README also claims support for open-source models running locally or in a data center, multiple installed models, model switching, and conversations using multiple models. Specific model names, versions, embedding models, and compatibility constraints are not stated in the repository metadata.

License and risk notes

The repository metadata reports SPDX value NOASSERTION and provides no license URL. The README displays both MIT and Apache license badges, so the applicable license cannot be determined conclusively from the supplied record. Review the repository's license files before use or redistribution.

Release and maintenance

No release or version information is included in the supplied record. The repository was created on 2023-07-07, uses main as its default branch, is not archived, and the supplied metadata lists its latest push as 2026-08-03.

ClearML

clearml/clearml

★ 6.8KPython