🆕What's New
Building the best AI assistant for the developers
Last updated
Building the best AI assistant for the developers
Last updated
Bito Inc. (c) 2024
New feature
Improved usability and formatting: We’ve redesigned the AI Code Review Agent’s comments in pull requests to make them more user-friendly and easier to read. With clearer formatting, you can now quickly spot key feedback and take action faster.
Skip redundant automated reviews: The Agent now skips automated full reviews if one is already running, saving valuable resources and cutting unnecessary costs. Manual reviews remain unaffected, giving you full control over the review process.
New feature
Enhanced code analysis for quality suggestions: We've made significant updates to the AI Code Review Agent's prompts for deeper code analysis, delivering relevant and actionable suggestions to help you improve your code quality.
Incremental review control at workspace level: You can now enable or disable incremental reviews at the workspace level, giving you greater flexibility to manage review settings across different projects. Please contact support@bito.ai to configure this feature.
New feature
Auto-review draft PR when marked as ready: The AI Code Review Agent now automatically performs a full or incremental review when a pull request moves from draft to ready status. This feature offers immediate insights to help you spot issues early, saving you time and streamlining your workflow.
Enable or disable incremental reviews: Incremental review can now be enabled or disabled at the Agent instance level, giving your team more control over the review process. Please contact support@bito.ai to customize this feature according to your team's needs.
Improvement and Bug fixes
Improved code review quality: We've optimized the AI Code Review Agent's prompts to provide more accurate, high-quality feedback, helping you identify and resolve issues faster.
New feature
Review committed code in your IDE: AI Code Review Agent can now review the latest commit, specific commits, or a range of commits, giving you better control over your code review process. This allows you to spot issues faster and ensure higher code quality before pushing to production.
Try the new commands in Bito extensions for VS Code and JetBrains:
/review #lastcommit
: Review the latest commit in the branch.
/review #commit=<commithash>
: Review a specific commit by its hash.
/review #commit=<commithash-1>..<commithash-2>
: Review a range of commits between two commit hashes.
Easier access to AI code reviews: You can now effortlessly invoke the AI Code Review Agent directly from the context menu by right-clicking in the code editor and selecting commands under the "Bito Code Review Agent" menu. This provides faster, on-the-go access to code reviews right where you work.
Quickly switch between Code Review and AI Chat: With the new "Back to AI Chat" button above the chatbox, you can easily jump back to AI Chat for conversations without interrupting your code review flow. Stay productive and manage both tasks seamlessly.
Improvement and Bug fixes
15% faster AI code reviews: We've optimized our AI Code Review Agent, cutting execution time by 15%. This improvement boosts your team's productivity, enabling quicker code reviews and helping you deliver high-quality features faster. Stay agile and keep your development process running smoothly!
New feature
Introducing incremental code reviews: Our AI Code Review Agent now automatically reviews only the recent changes each time you push new commits to a pull request. This saves time and reduces costs by avoiding unnecessary re-reviews of all files.
New feature
Introducing Changelist: a new tabular view in the AI Code Review Agent's feedback that displays key changes in a pull request, making it easy to spot important updates at a glance without reviewing every detail. Changelist categorizes modifications and highlights impacted files, giving you a quick, comprehensive summary of what has changed.
New feature
Introducing the Chain of Thought (CoT) approach: Our AI Code Review Agent now leverages the Chain of Thought (CoT) approach, which involves step-by-step reasoning to analyze your code more effectively. You will now receive more precise feedback, detailed explanations, and better solutions to enhance your code quality.
New feature
Introducing Claude Sonnet 3.5: Our AI Code Review Agent now exclusively uses the Anthropic Claude Sonnet 3.5 model, delivering more accurate and reliable code reviews. This new AI model offers superior performance in understanding complex code patterns and providing actionable recommendations.
New feature
AI code reviews with advanced code understanding: AI Code Review Agent within Bito’s extension for VS Code and JetBrains now offers deeper insights by fully understanding the entire repository context. This upgrade enhances code reviews, making them more relevant to your project’s needs and delivering actionable, precise suggestions directly within your coding environment.
To get started, configure the default @codereview agent instance with your Git information or use an already configured Agent instance from your Bito Cloud workspace.
We’ve made major upgrades to the 10X Developer Plan, including:
50% more GPT-4o requests - from 400 to 600 per month.
70% lower cost for additional GPT-4o requests - Reduced from $0.10 to $0.03 per request.
All at the same affordable cost of $15.
New feature
Support for Bitbucket: Now seamlessly integrate Bito’s AI Code Review Agent into your Bitbucket workflow. No downloads required—just a single-click install. Get AI-powered reviews posted as comments directly within your pull requests. Currently available on Bito cloud or through a self-hosted service via CLI.
New feature
Introducing GPT-4o mini: GPT-4o mini is now the default AI model for all basic AI requests. Available for both the Free Plan and the 10X Developer Plan, this upgrade brings improved performance and efficiency to your everyday coding tasks.
Introducing AI Chat Analytics: Explore our new analytics dashboard to gain insights into the AI Chat usage across your team. Track metrics such as total chats, chats per user, model usage, and more. Available under both the Free Plan and the 10X Developer Plan, this dashboard will help you make data-driven decisions for better collaboration and productivity.
New feature
Introducing the AI Code Review Agent for IDEs: Get human-like feedback on your code changes as you develop. Bito's AI Code Review Agent now supports VS Code and all JetBrains IDEs, including IntelliJ IDEA, PyCharm, WebStorm, and more. Developers can now address code issues early, accelerating development cycles and boosting team productivity.
The Agent is ready to use immediately without any setup required. Upgrade to the Bito 10X Developer Plan today and experience the future of code review firsthand.
New feature
Introducing code review analytics: Explore our new analytics dashboards for in-depth insights into your code review process. These user-friendly reports help you track key metrics such as pull requests reviewed, issues found, lines of code reviewed, and understand individual contributions.
Provide feedback on code review quality: For GitLab, we've added support for emoji-based reactions, while for GitHub, you can respond to feedback question(s) given at the end of code review comments.
This will help your team track the quality of suggestions provided by the AI Code Review Agent through the "Acceptance Rate" statistics available on the code review analytics dashboard.
Improvement and bug fixes
Track your Bito AI usage: Stay on top of how your team is using Bito AI. With the Requests Usage dashboard, you can easily see how many AI requests your team members and AI Agents are making. The Usage Management dashboard allows you to manage resources and control costs efficiently. This way, you can ensure that you're getting the most out of Bito AI without overspending.
Enhanced context understanding with prompt engineering: Our improved AI prompts now offer superior comprehension of your code context, significantly increasing the accuracy in identifying and addressing code issues. With Bito AI, critical problems are less likely to be overlooked, ensuring you receive a thorough and reliable code analysis.
Enhanced understanding of CSS, SCSS, and Terraform: Bito AI now analyzes both immediate context and broader structure of code in CSS, SCSS, and Terraform files, including nested configurations and dependencies. This enables Bito to provide more accurate suggestions, optimizations, and error detections.
For CSS and SCSS, Bito AI identifies patterns and offers improvements in style management and consistency. For Terraform, it understands and optimizes infrastructure as code, ensuring more efficient and scalable deployments.
Enhanced presentation of code review comments: We've made changes to how the AI Code Review Agent presents its feedback to enhance clarity and usability.
Improvement and Bug fixes
Prioritized Results from Open Files in IDE: When you ask a question about your code in Bito AI Chat, the AI now prioritizes answers from the files you have open in your IDE. By analyzing the relevant code first, Bito ensures more accurate and context-specific responses. This means you will now see relevant code snippets from what you are currently working on, saving you time and effort in finding the information you need.
Bito filters active files to exclude binary files, files with invalid extensions, files outside the project directory, and files in Diff view. The filtered list is then used to retrieve answers from the index.
New Feature
Filter Your Code Reviews, Exclude What You Don't Need: Introducing filters that give you control over what the AI Code Review Agent analyzes. They let you exclude specific files, folders, Git branches, or even draft pull requests from AI code reviews using glob/regex patterns.
For more information, see Excluding Files, Folders, or Branches with Filters.
AI Code Reviews with GPT-4o: GPT-4o integration brings a major upgrade to the AI Code Review Agent, enhancing code review quality, accelerating processing, and reducing costs.
Redesigned Comment Layout: We've redesigned the Agent's output comments to enhance readability and usability. The new layout is more organized and user-friendly, making it easier to navigate and understand the feedback. Important details are prioritized, saving you time and streamlining the review process.
Improvement and Bug fixes
Improved Code Review Quality with Less Noise: The Agent now analyzes each issue individually. This allows it to provide more focused and insightful suggestions tailored to the specific problem at hand. The cleanup process is also improved to reduce noise and unnecessary information. This means you'll get clearer, more actionable feedback that's easier to understand and implement.
New Feature
No Duplicate Comments: Bito's AI Code Review Agent now avoids posting duplicate comments. It achieves this by reviewing all its previous suggestions on a pull request before adding new ones. If it finds an existing suggestion that matches a new one, it skips the new one to avoid duplicates. This ensures the Agent provides only unique and relevant feedback on each pull request. However, if a user deletes a suggestion, the Agent will recommend it again in the next review.
More Relevant Suggestions: We've improved the accuracy and relevance of code suggestions by better understanding the context of your code.
Targeted Code Analysis via GitHub Actions: The AI Code Review Agent running via GitHub Actions now offers targeted code analysis with specialized commands! You can perform code reviews on specific areas of your codebase, including security, performance, scalability, code structure, and optimization.
/review security
: Analyzes code to identify security vulnerabilities and ensure secure coding practices.
/review performance
: Evaluates code for performance issues, identifying slow or resource-heavy areas.
/review scalability
: Assesses the code's ability to handle increased usage and scale effectively.
/review codeorg
: Scans for readability and maintainability, promoting clear and efficient code organization.
/review codeoptimize
: Identifies optimization opportunities to enhance code efficiency and reduce resource usage.
For more details, refer to Available Commands.
GPT-4o for Advanced AI Model Requests: We're thrilled to announce that Bito now leverages the power of GPT-4o for advanced AI model requests. This upgrade significantly enhances Bito's capabilities, providing you with faster responses. No need to update your IDE, CLI, or agents! This update works seamlessly behind the scenes, so you can experience the benefits of GPT-4o right away.
New Feature
5 New Commands for Targeted Code Analysis: The AI Code Review Agent now offers specialized commands designed to provide detailed insights into specific areas of your source code, including security, performance, scalability, code structure, and optimization.
/review security
: Analyzes code to identify security vulnerabilities and ensure secure coding practices.
/review performance
: Evaluates code for performance issues, identifying slow or resource-heavy areas.
/review scalability
: Assesses the code's ability to handle increased usage and scale effectively.
/review codeorg
: Scans for readability and maintainability, promoting clear and efficient code organization.
/review codeoptimize
: Identifies optimization opportunities to enhance code efficiency and reduce resource usage.
For more details, refer to Available Commands.
Improvement and Bug fixes
Limiting Advanced AI Requests for Agent Execution: Each execution of the AI Code Review Agent consumes advanced AI requests, which are now capped at a maximum of 20. This ensures a consistent experience and prevents users from unintentionally exceeding their advanced AI request quotas.
Combining Multiple Suggestions for Same Line and Range: We've improved how the AI Code Review Agent handles multiple suggestions for the same line or line range. Previously, each suggestion appeared as a separate comment, making it cumbersome for users to navigate through multiple entries. Now, all suggestions for the same line or line range are consolidated into a single comment, enhancing user convenience.
Improvement and Bug fixes
Handle Larger Pull Requests of Nearly Unlimited Size: No more limitations on pull request size! The AI Code Review Agent can now analyze more code at once, enabling smoother code reviews for pull requests of nearly unlimited size (subject to our Fair Use Policy).
More Focused and Relevant Recommendations: We've significantly enhanced the attention span of our AI Code Review Agent by optimizing its context window. The agent also filters out unnecessary suggestions (such as document notes, non-impacting changes etc.), delivering high-quality and precise code reviews that save you valuable time.
Static Code Analysis for Python via GitHub Actions: The AI Code Review Agent, running through GitHub Actions, can now analyze Python code using Astral Ruff and Mypy, in addition to its existing support for Java, Objective C, and C/C++. This enhances code quality and speeds up development cycles by proactively identifying errors and vulnerabilities in your Python codebase.
Improvement and Bug fixes
Improved Code Review Quality with Less Noise: We've enhanced the AI Code Review Agent to ensure you only see relevant suggestions. This means you'll no longer see suggestions that are already implemented in the code or those that primarily recommend documentation updates. Plus, each suggestion now displays the most relevant code changes (diff), making it easier to understand the context of the suggestion.
Better Code Review Process: The Agent now triggers automatic code reviews only for essential events, such as opening a pull/merge request. This reduces unnecessary reviews triggered by actions like closing and reopening. You can still request a manual code review by simply posting /review
in a comment on the pull/merge request.
New Feature
AI that Understands Your Code: The AI Code Review Agent can now better understand code changes in pull requests. It achieves this by analyzing relevant context from your entire repository, resulting in more accurate and helpful code reviews.
To comprehend your code and its dependencies, we use Symbol Indexing, Abstract Syntax Trees (AST), and Embeddings. For more information, see How does Bito’s “AI that understands your code” work?
Supported languages include JavaScript, TypeScript, Java, Go, Python, C, C++, PHP, and C#.
Static Code Analysis for Python: The AI Code Review Agent now actively supports static code analysis for Python code using Astral Ruff and Mypy, in addition to its existing support for Java, Objective C, and C/C++. It proactively identifies potential errors and vulnerabilities within your Python codebase, thereby improving code quality and accelerating development cycles.
Inline Comments: We've added support for inline comments in both GitHub and GitLab. This means code suggestions are now provided directly within each file, under the code diff. This offers a clearer view of the specific lines that need fixing.
Inline commenting is by default enabled for all new and existing AI Code Review Agent instances. The “Overview” section now shows the number of files changed and suggestions made in a pull request, along with a link that directs you to the specific file changes.
To manually trigger inline code reviews with comments directly on the code diffs, post the /review
command as comment on the pull request. You can also add a new optional parameter like /review #inline_comment=False
. This lets you receive a full-post code review instead of separate inline comments under the diffs.
New Feature
Added Support for GitHub/GitLab Enterprise (self-hosted): You can now integrate Bito's AI Code Review Agent with your on-premises GitHub or GitLab environment to enhance security. This is crucial for organizations in regulated industries with strict data compliance requirements.
Multiple Specialized Engineers: We've made significant improvements to the way the AI Code Review Agent analyzes your code. It now acts as a set of specialized engineers each analyzing different aspects of your PR. They analyze aspects such as Performance, Code Structure, Security, Optimization, and Scalability. By combining and filtering the results, the Agent can provide you with much more detailed and insightful code reviews, bringing you a better quality code review and helping you save time.
New Feature
Light and Dark Themes: In Visual Studio Code and JetBrains IDEs, you can choose between a light or dark theme for the Bito panel to match your coding environment preference. For VS Code users, Bito also offers an adaptive theme mode in which the Bito panel and font colors automatically adjust based on your selected VS Code theme, creating a seamless visual experience.
The IDE customization settings are accessible through the new toolbar dropdown menu titled "Extension Settings".
Font Size Control: Take control of your code readability! Within the Bito extension settings, you can now adjust the font size for a comfortable viewing experience.
AI Smartly Identifies Your Question's Intent: Bito automatically figures out if you're asking about something in your code. If it's confident, it grabs the relevant parts of your code from our index and feeds them to the Large Language Models (LLMs) for accurate answers. But if it's unsure, Bito will ask you to confirm before proceeding.
Don't worry, existing cue words such as "my code" and "my repo" still work perfectly!
Improvement and Bug fixes
Fixed Multi-Level Paths in .bitoignore: The .bitoignore
file now correctly handles paths with multiple subdirectories. Previously, using patterns to ignore deeply nested directories wasn't working as expected in VS Code. For instance, if you add a pattern like project/data/subfolder
to your .bitoignore
file, it will now correctly ignore all files and subdirectories within the project/data/subfolder
directory.
Fixed Slash Command Usability Issue: Bito chatbox no longer blocks users from typing content that starts with a slash /
, such as /bin/bash
. The Template panel now disappears entirely if no matching templates are found after entering additional characters following the slash.
Improvement and Bug fixes
400 Advanced AI Requests Monthly: Exciting news from Bito's 10X Developer Plan! We've raised your Advanced AI Models' quota from 100 to 400 requests per month. This significant increase allows for uninterrupted usage, giving you the freedom to explore more without worrying about limits. For more information, please visit our Pricing Page.
Advanced AI Default in Bito Chat: 10X Developer Plan now starts with Advanced AI Models as default for Bito AI Chat in the IDE extension, ensuring immediate access to top-tier coding assistance for enhanced efficiency and creativity.
Suggest Code Inside Brackets: We've resolved the issue that prevented AI Code Completions within inline brackets, including arrays, function headers, and more. Now, when you input code within brackets—like creating an array in PHP, such as $numbers = [89, 7, 22,<Bito_DISPLAYS_CODE_COMPLETION_HERE>];
—right after you add a space or comma following the last array element, Bito immediately provides suggestions to help you fill in the code faster. It's not just for arrays; this works for all kinds of brackets in your code.
New Feature
Bito's AI Code Review Agent: Your team's new automated code reviewer! Leveraging the best Large Language Models such as GPT-4 and Claude 2.1, this tool is here to transform how senior developers review code.
Here’s what it offers:
Automated Code Reviews: Instantly spots bugs, code smells, and security vulnerabilities in your Pull/Merge Requests.
Integration with Git Providers: Works with GitHub, GitLab, and Bitbucket (coming soon), posting recommendations as comments right within the Pull Request.
Comprehensive Analysis: Utilizes static analysis and OSS vulnerability tools for thorough insights.
Tailored Suggestions: Offers specific, actionable advice to improve your code, ensuring it’s up to standard.
Privacy First: Prioritizes your security and confidentiality; no code is read or stored.
This AI Agent not only speeds up code merging but also elevates code quality, letting you focus on what matters. Ready to code smarter, not harder? Welcome to the future with Bito.
Experience this cutting-edge tool on Bito Cloud or opt for a self-hosted service via CLI, webhooks, or GitHub Actions.
It's included at no extra cost with Bito's 10X Developer plan. For more information, please visit our Pricing Page.
Improvement and Bug fixes
DLL Files Excluded From Index: To improve the indexing speed of your projects, we will now be excluding DLL files by default.
Fixed localStorage is Full Error: Some of you might have bumped into a bit of trouble with a JavaScript error saying your localStorage is full. We have resolved that issue. So, no more storage woes!
Fixed AI Code Completions Setting: We have fixed the Java exception that occurs when you try to switch the AI Code Completions feature on or off without an internet connection.
Fixed "Enable/Disable This" Link: In Bito panel, we've fixed the "Enable/Disable This" link that wasn't working properly before. Now, you can use it to access the AI Code Completions settings without any trouble.
Improvement and Bug fixes
Fixed Backward Compatibility Issue: Bito now works seamlessly with JetBrains version 2021.x and onwards. Thanks to your valuable feedback, we have identified and fixed the issue that was breaking backward compatibility of the Bito extension in JetBrains IDEs, such as IntelliJ IDEA.
If you encountered annoying error messages like Cannot load class co.bito.intellij.webview.BitoWindowFactory
, they're a thing of the past now! Simply update your Bito extension to the new version 1.2.4, and you'll be all set.
New Feature
AI Code Completions Now in JetBrains IDEs! Say goodbye to long hours of staring at your screen trying to figure out the next line of code. With Bito, as soon as you start typing a few characters or jot down a comment, it will start offering real-time, tailored suggestions that fit your coding style.
Powered by the latest and best-in-class LLMs, Bito offers autocomplete capabilities for SQL queries, regex patterns, functions, loops, if-else blocks, and more—right within your IDE as you code
Curious to know more? Dive into our documentation to see how this feature can make your coding life a breeze.
AI Code Completions are disabled by default. Learn how to Enable/Disable AI Code Completions in settings.
Slash /
command support for Templates: Using templates in Bito just got a whole lot easier and cooler! All you have to do is type a forward slash /
right at the start in the Bito chat box. Once you do, the template menu will open from where you can quickly select and use the template you want.
Want to narrow down your choices? Simply start typing after the /
slash, and it'll only show you templates that match your words. And hey, you can also use the arrow keys, or Tab and Shift + Tab, to navigate the templates menu.
Give it a try and make your coding experience super smooth and fun!
Currently, the AI Code Completions feature is only available in Bito's VS Code extension. It will be coming soon for JetBrains IDEs.
New Feature
AI Code Completions: Get real-time and personalized code completions from AI that understands your code. Just type some initial code in your file or provide explicit instructions as comments, and watch Bito suggest the code you're most likely to write next.
Powered by the latest and best-in-class LLMs, Bito offers autocomplete capabilities for SQL queries, regex patterns, functions, loops, if-else blocks, and more—right within your IDE as you code
Curious to know more? Dive into our documentation to see how this feature can make your coding life a breeze.
In release 1.2.0, AI Code Completions are disabled by default. Learn how to Enable/Disable AI Code Completions in VS Code settings.
Improvement and Bug fixes
Fresh Look: The Bito Panel UI is now more intuitive and pleasing to work with. We've put most buttons closer to where you type, and guess what? You've got 30% more screen space to see the AI's answers!
Custom Prompt Templates Output: Select where the results from your Custom Prompt Templates should be displayed. Your options are:
Display in Bito panel (Default)
Output in diff view
Clickable File Links: Source code file names that are mentioned in AI Chat answers are now clickable. A single click will open the respective file directly in your IDE. No more hunting for files!
Speed Boost: We've supercharged indexing in VS Code to match pace with JetBrains IDEs (e.g., IntelliJ IDEA, PyCharm, etc.) and optimized how we calculate index sizes for both.
Bug Fixes: No more hiccups with the Generate Comment feature in Bito menu; it will now correctly show outputs in Diff View. Also, we have resolved some customer reported issues where Diff View wasn't working.
Improvement and Bug fixes
Performance Upgrades: Bito can now process multiple chunks at once, slashing network calls and giving a considerable boost to indexing speed for both JetBrains & VS Code users. Plus, VS Code users can expect an even swifter performance with our newly integrated batched writes.
Bug Fixes: Our team has diligently tackled issues from V1.1.7:
We've resolved the race conditions that occurred when switching between multiple IDE windows.
In Bito's "AI that Understands Your Code" feature, we've corrected problems with the "my code" keyword in Spanish and Portuguese. Additionally, Polish language keywords are now supported.
The keyboard shortcut for code insertion in the IDE extension has been fixed.
Improvement and Bug fixes
Improvements in AI that Understands Your Code
After introducing code-understanding capabilities in our previous versions, we've now made it even better in this release. Learn more about Bito's AI that Understands Your Code feature.
Introducing .bitoignore: You can now tell Bito to skip certain files and folders during indexing. Perfect for indexing those big repositories without the unimportant stuff.
30% to 40% Faster Indexing: We have improved the way your repo is indexed. Now, Bito can understand every 10MB of code in just 16 to 18 minutes, a big leap from the old 30-minute mark. Enjoy the speed!
Up to 120MB Max Repo Size: You can now index larger repositories. We've bumped up the indexing limit to 120MB per repo. Need to trim a bit? Use the .bitoignore feature to exclude specific files and folders and remain within the limit. Happy indexing!
Over-100% Indexing Bug Fixed: Good news! We've fixed that quirky bug where the indexing went beyond 100% during repo reindexing. No more overachieving percentages!
Billing Related Bug Fixes
Billing Invite Glitch: All Sorted! Had trouble inviting users via the Bito Web UI for billing? Good news! We've fixed that hiccup. Invite away!
Improvement and Bug fixes
File/Folder Filtering Improved for Indexing: We've added more files and folders to the exclusion list, speeding up the indexing process by preventing Bito from indexing unnecessary files and folders.
Progress Indicator for Indexing: Index building now has a progress indicator based on the number of files indexed. The progress indicator for the current folder is updated in real-time. However, for projects in the "Other projects" section, the progress indicator is updated every 5 minutes.
Added "Indexing is paused" status: This status will be displayed for a project that was previously being indexed but is now paused for some reason. Generally, if you close the IDE while the project is being indexed, its status will change from "Indexing in progress" to "Indexing is paused."
New Feature
AI that Understands Your Code: Bito's most requested feature is here. Bito AI now comprehends your codebase inside out using AI indexing, ensuring privacy as the index stays on your system.
Simply add "my code" or “我的仓库” in Chinese to your AI requests and watch Bito efficiently navigate your repository, providing precise solutions.
Try it out with queries like:
Find errors in scraper.py in my code.
Write frontend and backend code for user authentication, using the authentication service in my code.
List files and changes needed to add a "desc" column in the "raw_data" table in the "dailyReport" DB in my code.
Have an even better experience with Bito’s AI assistant, seamlessly attuned to your codebase. Explore more in the documentation here.
This feature is currently available to our paid plan subscribers and will be launched for free plan subscribers soon.
Improvement
Expanded Contextual Memory: Bito now has a larger contextual memory and output capability, that can process up to 40,000 characters (around 18 single-spaced pages). This expanded memory encompasses the provided prompt, existing context from previous chats, and the generated output.
Our latest version of Bito offers 3X longer conversation memory, ensuring contextually accurate responses. Chat like never before!
New Features
Language support for AI Output: We're excited to announce that Bito now supports multiple languages in IDEs, allowing you to converse in your preferred language. With this update, you can easily switch to your desired language within the chat interface.
Supported languages: English, Bulgarian, Chinese (Simplified), Chinese (Traditional), Czech, French, German, Hungarian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Turkish, and Vietnamese.
How to Access:
Click on the settings icon in the plugin interface's top right corner.
Select your preferred language from the list of supported languages.
Save your choice; Bito will now communicate with you in the selected language.
Note: All responses from Bito will appear in the selected language, regardless of the input language.
Enjoy the convenience of conversing with Bito in your native language and take your coding experience to a new level!
Custom Templates in IDE Context Menu: Your requests are our priority! We are pleased to introduce running "Custom Prompt Templates" to the IDE context menu. Building on the success of our previous release, where we introduced custom templates in the Bito plugin, we've taken it a step further. Now, you can access your personalized templates directly from the context menu in your IDE from "Run Custom Prompt Template"
Clicking "Run Custom Prompt Template" opens up a list of your templates in the Command Pallet or other secondary menu, depending on your IDE.
Introducing Context in Bito CLI: Bito CLI can now remember previous conversations, allowing for a more seamless and contextual experience. With the addition of context file support, you can now pass a context file using the -c command to preserve the context and history of your conversation. Watch our helpful tutorial to learn more about utilizing context files in Bito CLI.
Say goodbye to repetitive context-setting and dive straight into productive and context-aware discussions with Bito CLI! 💬💡
Improvement and Bug fixes
Bito Version 1.0.135 and IntelliJ IDEA 2021.1.x Compatibility: We recently discovered an issue where Bito version 1.0.135 was not compatible with specific older versions of IntelliJ IDEA. We want to assure you that we have addressed this compatibility concern, and Bito is now fully compatible with IntelliJ IDEA.
Automatic CLI updates: Say goodbye to manual updates! We're thrilled to introduce the CLI Auto-update feature. With this, Bito CLI will automatically check for updates whenever you open it. If a new version is available, the CLI will seamlessly update itself, ensuring you're always equipped with the latest enhancements, bug fixes, and features. Stay ahead of the curve without the hassle of manual updates!
New Features
Introducing Custom Prompt Template: Bito allows you to create and save custom prompt templates for IDEs. Define a template name and prompt, and Bito will execute it on the selected code. You can add up to 4 custom templates and edit or remove them as needed. Learn more about Custom Prompt Template.
Improvement and Bug fixes
Resolved OTP verification issue: Unfortunately, some customers experienced an issue last week where they could not complete OTP verification during login to the Bito Plugin. We are happy to report that this issue has been resolved in the latest release.
Fix for Context Menu and Shortcuts on Ubuntu 2023.1 JetBrains: Fixed an issue where the context menu and shortcuts were not functioning correctly on Ubuntu 2023.1 JetBrains, causing the selection of code and options like explain code not to work as intended. This has now been resolved.
CLI "My Prompt": We're excited to announce the launch of Bito CLI My Prompt, our first step in empowering developers to automate tasks with AI. My Prompt makes creating accurate and concise prompts for your development tasks easy, reusing them with simple Bito commands. It allows you to automate various coding tasks, such as generating commit messages, test data, or code documentation. The possibilities are limitless. Two new command line options support My Prompt feature:
-p option: To specify a file containing the prompt or instructions for the AI models to operate.
-f option: To specify the file to perform the operation/instructions described in the prompt file.
Here are two examples for you to see My Prompt in action:
Create Git Commit Messages and Markdown Documentation with Ease using Bito CLI My Prompt
Generate test data using Bito CLI My Prompt:
Generate Unit Test Shortcut (Waitlist): With this new shortcut button, developers can now easily generate unit tests for their code with just a click. This new feature is designed to streamline the testing process and improve code quality, making it easier for developers to deliver high-quality work.
Be the first to try our exciting new feature! You can now add yourself to the waitlist by clicking on the icon.
Improvement and Bug fixes:
Improved Explain Code Shortcut: Previously, the Explain Code shortcut button provided a summary of the code logic, but with our latest update, it now also provides a detailed, step-by-step explanation of the code. We have merged two previous shortcuts into one. Now, developers can easily understand the code logic and its components.
We've fixed an issue where the CTRL+C command was not working on Mac OS. Now, you can easily copy and paste text without any problems.
We've resolved an issue with context menus and shortcuts not working on Ubuntu 2023.1 JetBrains. This fix ensures that all shortcuts and context menus work as expected, making it easier for developers to navigate the IDE.
We've fixed an issue on Linux version 2023.1 IntelliJ where users could not type anything in the text area of our chat interface. This fix ensures that users can input text as expected.
Chat Session History: In the last release, we added Context-Aware chat. This release lets you view past chat sessions on the chat interface. The history feature is located at the top right corner of the chat interface. With this, users can easily revisit their past conversations, making it easier to pick up where they left off. Additionally, users can share their past chat sessions with colleagues, making it easier to collaborate and share knowledge. The new chat session History adds another layer of convenience and functionality to the AI assistant.
GPT-4: We are thrilled to announce that we have integrated GPT-4, the industry's latest and most powerful generation model. With GPT-4 integration, Bito's AI assistant is now even more powerful and capable of providing detailed and accurate responses to various programming-related queries. Developers can ask complex questions and receive natural language responses that are both helpful and informative.
Long Code Input: Bito fully leverages GPT-4's increased token limit. With this, users could input and receive feedback on larger blocks of code, making it easier to troubleshoot and optimize code for performance. The exact size of the code depends on many factors, including how long the current chat session is. Bito can take 2-3x longer code input than the input length supported by the previous version of Bito.
Note that we use a combination of GPT-4 and GPT Turbo 3.5 models. We route long, complex prompts to GPT-4 and short prompts to Turbo 3.5 models. This helps us balance the quality of output that Bito generates and the cost of servicing users. If you ask Bito about the model in use, most likely, it will say GPT-3, as this simple prompt is routed to Turbo 3.5.
We may offer a paid version of Bito in the future that always uses GPT-4 or other advanced models. However, our internal testing has shown that the output quality for simple and short prompts is not very different between 3.5 and 4. The full benefit of GPT-4 is realized in complex, long prompts or when prompts are related to niche facts that models are likely to hallucinate. We have not seen "hallucination" as a major issue for the development-related use cases that Bito is targeting.
Improvement and Bug fixes:
Stop Generation in Chat Interface: We have added a Stop Generation button to our chat interface, allowing users to stop the response generation process anytime. This feature will be especially helpful in cases where users have asked a question incorrectly or want to modify their question before receiving a response.
Diff View Bug Fix – We have resolved a bug in our VS Code Extension v1.0.132. The bug affected one of our key features, Diff view, which provides a side-by-side view of changes proposed by Bito AI in the IDE. The bug was causing multiple diff views to open simultaneously, which could be confusing. This has been fixed now.
Context-Aware Chat: Our latest update is the most critical one since our launch. Many of our users had requested an interactive chat experience similar to ChatGPT, which was unavailable in earlier versions. With this new update, you can now ask follow-up questions to refine the output, which considers the chat history for context. You can initiate a chat session with a custom request, use Bito Shortcuts like Performance Check, and follow up with additional prompts. To give you an idea of how it works, here is an example of an interactive chat session.
Instant Results: With the latest release of Bito, you can now experience instant results that appear within seconds. In previous versions, users had to wait patiently while being entertained by amusing messages, but we understand that this could become frustrating over time. This new update implements streaming output, allowing you to access the information you need with minimal delay immediately.
Note: Context-Aware Chat and Instant Results are available in the latest version of Bito CLI and Chrome Extension.
Improvements & Bug Fixes
Dynamic Font Size & Family: This feature in Bito adjusts the interface's font size and family dynamically based on the font size you have set up in your Integrated Development Environment (IDE) settings. The goal of this feature is to make the chat interface more accessible and comfortable to use for all users, regardless of their font preferences, visual acuity, or screen size.
Configure Keyboard Shortcuts: This feature allows users to customize the default keyboard shortcuts in Bito to avoid conflicts with other IDE-specific shortcuts they may be using. By configuring keyboard shortcuts, users can streamline their workflow and increase their productivity by accessing Bito functions more quickly and efficiently. The instructions for configuring keyboard shortcuts in Bito are in this link.
Diff View: Bito Shortcuts added many quick operations on your existing code, such as checking performance, improving code readability, or security check. The Diff View in IDE now gives a side-by-side view of the changes proposed by Bito AI. Bito automatically creates a temp file with the proposed code changes and displays diff view with your current code file. IDE's built-in "diff review feature" lets you review and accept/reject the changes. We love that Bito AI can improve your code in many ways, but we want you to be in control while incorporating changes to your code repository. Diff view is built just for that.
Keyboard Accessibility: We have added keyboard accessibility to all UI elements in Bito. You can navigate with Tab and Shift + Tab. Standard keys such as Enter, ESC, and Up/Down arrows work. You can also move between Q&A with the Up and Down arrow keys once the Q/A container is focused.
Keyboard Shortcuts: Keyboard shortcuts for the most commonly used commands.
SHIFT+ CTRL + O: Brings Bito panel in focus. With this shortcut, you can toggle Bito Panel on/off in the JetBrains extensions. In VS Code, the shortcut will open the Bito panel (if not already opened). You can use shortcuts of other tabs (e.g. CTRL + SHIFT + E for Explorer) to close Bito.
SPACEBAR: Puts cursor in the chatbox when Bito panel is in focus.
ENTER: Execute the chat command.
CTRL+ ENTER or SHIFT + ENTER: Adds a new line in the chatbox.
The following shortcuts work on the Q&A when a Q&A is selected.
CTRL + C: Copy the answer to the clipboard.
CTRL + I: Insert the answer in the code editor.
CTRL + D: Toggle the diff view (when Diff View is applicable)
CTRL + UP/DOWN (Mac: CTRL + SHIFT + UP/DOWN): Expand and collapse the code block where applicable.
CTRL + L : Regnerate the answer.
CTRL + U: Modify the prompt for the selected Q&A. Bito copies the prompt in the chatbox that you can modify as needed.
CTRL + M: Modify the most recently executed prompt.
CTRL + UP/DOWN (CTRL + SHIFT + UP/DOWN on Mac): Expands and collapses the "Shortcut" panel in Bito.
Sharing: You can easily share a single question and answer to any channel. Every question and answer has a web permalink that you can share on Twitter and E-mail directly. Copy and paste the link into Slack, Teams, Jira, or Confluence.
Regenerate the results of any prompt with a single click.
Edit prompt lets you modify a previously executed prompt. You can edit any prompt with the "Edit Prompt" action next to the question.
To modify the most recent prompt, click the "Redo" icon in the chatbox.
Improvements & Bug Fixes
The Bito icon in the JetBrains IDE is optimized, so the extension panel doesn't take up space when not in use.
Code Block formatting and syntax highlighting make it easy to distinguish code from the text content in the chat output. When you use Shortcuts to transform any existing code, the original code is collapsed by default in the question box, thus giving more space for the answer, and the output code is formatted and syntax highlighted.
Several UI improvements to make Bito look and work great in your IDE. We have reduced padding around the questions and answers blocks so you see more content. The scrollbars are optimized to take up less space and reduce visual clutter.
Shortcuts give your existing code power of Bito AI. Select any code snippet in your code editor, and use Bito Shortcuts to run security, performance, or style check. Know how the code works, what it does, or generate comments. You can clean code, improve readability or automatically add exception handling. Behind the scene, we automatically generate the best possible prompt for the best AI output, so you don't have to spend time with prompts.
Fixed crash when using Bito JetBrain extension with PyCharm.
Fixed Git dependency issue in the Visual Studio Code extension.
Additional pre-defined prompts to improve the quality and readability of the code.
Improve Readability - Makes code more readable by organizing the better
Clean Code - Remove debugging, commented code, and log-related code.
Make Robust - Add exception/error handling to the code.
Add Steps - Add explanation steps to the block of codes.
Improved UX for smaller screen resolution.
Pre-defined prompts for various checks on your existing code. Select your code and choose one of the below options:
Explain Steps: Explain "How the code works?".
Explain Code: Explain "What does this code do?"
Generate Comment: Generate a comment for the selected code.
Performance Check: Check the selected code for performance and gives pointers to fix the performance. It also rewrites the code with the suggested performance fixes,
Security Check: Check the selected code for security issues and gives pointers to fix them. It also rewrites the code with the suggested security fixes,
Style Check: Check the selected code for the style issues and gives suggestions to fix them. It also rewrites the code with the suggested style fixes.
Command Pallet support in Visual Studio Code. All Bito commands are accessible in the Visual Studio Command Pallet.
Insert output from Bito AI into your code file with a single click. This is great for inserting any generated code in your code editor. It inserts the code wherever your cursor is, or if you want to replace the existing code, select the code and insert the new code.
Keyboard support for the AI Chatbox. Send the request with 'Enter' and add a new line to the chat with 'CTRL+Enter.'