The Future of AI Chat is Here
Whether you're a developer, creative professional, researcher, or just someone who loves having intelligent conversations, v5.1.0 brings powerful new capabilities that make your AI assistant more helpful, faster, and smarter than ever before.
This isn't just an update; it's a complete evolution. From connecting directly to your GitHub repositories to remembering your unique preferences, LeemerChat is now personalized to you and your code.
Everything Shipping in v5.1.0 (Parts 1–7)
- Tool system re-architecture and utility tool expansions.
- Image generation access fixes plus Mistral model support.
- Major chat latency reductions and faster response times.
- Concurrent chat generation with background job support.
- AI Memory with opt-in preference learning across chats.
- Second Thought expert consultations across top frontier models.
- GitHub Connect and codebase chat mode for instant repo context.
Codebase Chat: Talk to Your Code
Imagine asking your AI assistant: "How does authentication work in my codebase?" or "Where is the payment processing logic?" and getting instant, accurate answers with full context of your entire repository.
Codebase Chat Mode connects your GitHub repositories directly to LeemerChat. It's like having a senior developer who's memorized your entire codebase, ready to answer questions 24/7.
Why not GitHub Copilot?
LeemerChat gives you a wider model lineup than Copilot plus higher usage limits for the price. That means you can choose the right model for the job (fast, deep reasoning, or long-context) without running into caps as quickly. If you want repo-aware chat with flexible model choice, this is it.
We're also rolling out PowerCode agent connectivity soon. It uses the same GitHub access you already grant, so advanced coding agents can operate on your connected repositories without a second setup.
What You Can Do
- Ask Anything"How does the login flow work?"
- Instant File AccessRead and understand any file
- Smart ExplorationIntelligent two-stage search
- Persistent ContextStays active across chats
AI Memory: It Finally Listens
Tired of repeating "be more concise" or "I prefer technical explanations"? With AI Memory, your assistant learns your preferences and applies them automatically to every conversation.
Natural Learning
Just say "I prefer shorter answers" in conversation, and it's saved automatically (opt-in).
Topic Specific
Set different preferences for math, coding, or creative writing tasks.
Cross-Device Sync
Your preferences follow you everywhere, across all your devices.
Full Control
View, edit, or clear your memory at any time in Settings.
Lightning-Fast Responses
We've dramatically improved response speed. Your AI assistant now responds up to 3x faster in many cases, with sub-3-second response times becoming the norm for fast models.
We eliminated unnecessary delays, optimized tool usage, and streamlined the processing pipeline. The result is conversations that feel more natural, fluid, and responsive.
Multi-Task Like a Pro
Ever wanted to start a new conversation while waiting for a response? Now you can! Concurrent Chat Generation lets you multitask across multiple conversations without losing progress.
- Start new chats while others generate
- Visual tracking in sidebar
- Notifications for completion
- Seamless switching
Second Thought: Expert Consultations
Sometimes you need an expert's perspective. The Second Thought feature lets your AI assistant consult specialized expert models for verification, alternative perspectives, or deeper insights.
Alternative Frontier (ZAI GLM-4.7)
Creative problem-solving and alternative perspectives when you're stuck.
General Frontier (OpenAI GPT-5.1)
High-quality verification and deep expertise for complex topics.
Web-Enabled Fast (Perplexity Sonar)
Current information and rapid web research for up-to-date answers.
The Second Thought Pipeline
Visualizing how LeemerChat seamlessly consults experts in real-time.
Main Model
OrchestratorAnalyzes the user request and determines an expert is needed.
Tool Call
second_thoughtConstructs a targeted prompt and triggers the expert tool.
Expert Model
SpecialistIndependent model (e.g., GPT-5.1, Sonar or GLM-4.7) processes the sub-task.
Final Reply
SynthesisMain model integrates expert insights into a cohesive response.
New & Improved Tools
We've completely revamped our tool system. New utility tools include Unit Converter, Color Converter, QR Code Generator, Password Generator, and Hash Calculator. We also fixed currency and crypto tools, improved weather displays, and unlocked image generation for Mistral models.
A Note from the Founder
Repath "Ray" Khan, CEO & Founder
"We are smashing 2026 so far! We've already generated over 500 million tokens this year alone. The goal is to hit 1 billion before the end of January. This update is a huge step towards making LeemerChat the ultimate tool for builders."
Experience v5.1 Today
These features are live and ready to use. Connect your GitHub, set your preferences, and start chatting.
Launch LeemerChat