jan Releases
110 releases of janhq/jan
- v0.5.110.5.11
Platform support expansion for MacOS universal build combined with stability fixes for markdown rendering and health checks.
Dec 5, 2024
- v0.5.100.5.10
Enhanced privacy documentation and expanded model support, including new Groq and Anthropic models with increased token limits.
Dec 3, 2024
- v0.5.90.5.9
Focuses on backend configuration capabilities (Proxy, HF Token) and UI enhancements for code blocks and remote model handling.
Nov 22, 2024
- v0.5.80.5.8
This release integrates Cortex.cpp as a primary provider and introduces a new nightly app channel. It focuses on UI refinements including auto-focusing chat inputs and reducing blur effects, alongside various bug fixes.
Nov 13, 2024
- v0.5.70.5.7
The release adds support for markdown rendering in user messages and stabilizes thread instructions as a feature. It also includes updates for the beta version and fixes for multi-modal download states.
Oct 24, 2024
- v0.5.60.5.6
A maintenance release focused on improving error handling during the model import process to prevent crashes and improve stability.
Oct 4, 2024
- v0.5.50.5.5
Enhances the model selection UI with a new tabs component and introduces auto-resizing textareas. It addresses several issues related to model downloads, metadata, and context lengths.
Oct 2, 2024
- v0.5.40.5.4
Enables searching the model dropdown by configured models and adds setting options to the dropdown menu. It also addresses performance issues observed on CPU inference.
Sep 17, 2024
- v0.5.30.5.3
Integrates Cortex as a JavaScript module and introduces a new starter screen experience. It improves model settings preservation and manages cloud models from the threads screen.
Aug 29, 2024
- v0.5.20.5.2
Adds support for Claude 3.5 Sonnet and introduces spell check configuration for chat inputs. It also adds shortcuts for app logs and fixes various UI layout issues.
Jul 15, 2024
- v0.5.10.5.1
Updates the Cortex backend and integrates the NVIDIA API. It resolves issues with window controls on Linux and duplicate message roles.
Jun 21, 2024
- v0.4.140.4.14
Updates the Next.js framework to version 14.2.3 and Cortex to version 0.4.9. It focuses on Phi-3 long context updates and OpenAI endpoint mapping.
May 27, 2024
- v0.4.130.4.13
Replaces the Nitro engine with Cortex-cpp and adds support for OpenRouter and Anthropic extensions. It includes fixes for context length validation and deeplinking.
May 20, 2024
- v0.4.120.4.12
This release adds LaTeX support for math rendering, introduces Llama 3 and Groq models, and provides a Windows NSIS installer. It includes a refactor of the Hugging Face extension and fixes numerous bugs related to thread titles, model loading, and GPU settings.
Apr 25, 2024
- v0.4.110.4.11
Adds support for the Mistral inference engine and moves logs into the monitoring extension. Includes fixes for race conditions when switching models, better error handling for core, and handles CJK input methods correctly.
Apr 15, 2024
- v0.4.100.4.10
Features a revamped homepage and download page, along with the Groq Inference Extension. Includes TensorRT-LLM documentation and fixes model hub ID mismatches.
Apr 2, 2024
- v0.4.90.4.9
Introduces the Nitro-TensorRT-LLM Extension and fixes various regressions for tensorrt nightly builds. Improves documentation and fixes issues regarding Quick Ask feature toggles.
Mar 19, 2024
- v0.4.80.4.8
Integrates Turborepo for better build management and adds an import model feature. Fixes image upload, drag-and-drop support, and addresses issues with model import crashes.
Mar 11, 2024
- v0.4.70.4.7
Adds Vulkan support for AMD and Intel Arc GPUs, Playwright for testing, and a troubleshooting modal. Improves UI/UX for GPU acceleration and model recommendations.
Feb 26, 2024
- v0.4.60.4.6
Introduces a factory reset feature, snackbar component, and umami analytics. Adds an HTTP API to start/stop models and improves the UI for thread management.
Feb 5, 2024