This year will be the biggest vintage in a long time at Hex-Rays. We’re planning to deliver to our user community more improvements and capabilities than ever before, thanks in part to our growing team, productivity gains from dev tools, and some good old fashioned ambition.
Looking back at 2025
Before we look ahead at our 2026 goals, let’s take a look at how we closed out last year. In 2025, we publicly shared our Product Vision & Goals. With the three IDA releases during the year, much of that vision became a reality, along with several other advancements.
Highlights from the last 12 months
- Delivered IDA 9.1, 9.2, and 9.3
- Introduced the Domain API
- Rolled out hcli, the command line utility
- Open-sourced the IDA SDK
- Rebuilt and modernized the plugin repository
- Deployed the new plugin manager
- Released the enhanced license server, for customers sharing seats across a fixed number of users, i.e. floating licenses.
Bonus Capabilities & Plugins
We also shipped a few unplanned, but fun, additions:
Still in Progress
Some initiatives saw substantial growth but are not yet fully complete:
- FLIRT 2.0
- Live Collaboration feature for Teams, which we are continuing to refine for compatibility with your preferred live chat solutions.
Community Contributions
Last but certainly not least, we remain deeply grateful to our vibrant IDA user community:
- 92 new plugins contributed in 2025 (insane!)
- 80 existing plugins are now compatible with the new plugin manager
Thank you all for your invaluable contributions!
Looking ahead at IDA in 2026
Building on 2025’s foundation, here’s where we’re heading next. Below are the themes shaping IDA in 2026. As always, priorities may evolve and timelines may shift.
This is far from an exhaustive list. If there’s something high on your wish list that isn't here, let us know at product@hex-rays.com. Who knows, it may already be on our (secret) list too.
Expanding IDA’s Technical Depth
Modern language support: In the same vein as our 2025 Golang improvements (check out the 9.2 and 9.3 updates), we are bringing significant improvements to our decompilers, disassemblers and loaders for code written in Swift, Rust, C++ (including improved support for containers and templates), Python, and native .NET AOT code. Furthermore, our decompiler frontend has matured enough to support JVM-based languages (Java, Kotlin, …) in the future.
Improved mobile ecosystem support: Expect advances in iOS dyld shared cache analysis, numerous iOS kernel cache improvements, and expanded decompilation for Android, both in the separate ART and native worlds, but also (confusingly) the fused ART/Native AOT.
New decompilers: Targets for 2026 include Android™ Dalvik, Infineon TriCore™, Qualcomm Hexagon™, Andes AndeStar™ V3 and the remaining parts of AndeStar™ V5 not yet covered by our current RISC-V decompiler. In the pc world we‘re also improving SIMD/AVX handling, decompiler interactivity, the type system, and deeper analysis capabilities.
Additional processor modules / disassemblers: Several platforms are in the pipeline, including Qualcomm Hexagon™, Alibaba Hangzhou C-SKY, Loongson’s Loongarch, and further ARM SVE enhancements. At the same time we continue to work on our existing disassemblers, by expanding regtracker functionality, recovering more switch case patterns, and more thoroughly resolving indirections in general.
Firmware, loaders and file formats: Planned improvements include enhanced support for UEFI and Shannon/Qualcomm baseband, and a more robust (less finicky) experience when loading blobs into IDA. For mainstream platforms we plan to carve out more metadata from the file formats to provide richer context to analysts.
Performance & Experience
Performance, performance, performance: Loading PDB & DWARF files OOMing your machines? Waiting far too long for the dyld shared cache to load? Wanting IDA to work faster on large binaries? And much faster UI and headless IDA experiences? If so, expect steady improvements throughout 2026 and beyond.
UI/UX/API modernization: We are modernizing the user interface, expanding host platform support, simplifying APIs, introducing CI/CD templates for plugin developers, and deepening integrations with third-party sandboxes, debuggers and emulators.
Collaboration, Code Intelligence & Malware Analysis
Version control, version diffing, team collaboration: With the upcoming update to the Teams add-on, users will be able to connect to any Git-compatible server (e.g., Git, GitHub, GitLab) for version control. The add-on will offer integrated diffing and merging from within IDA, along with “deep links” to point to a particular code region and share them with team members.
Code recognition and signatures: FLIRT 2.0 will be shipped alongside IDA 9.4. In addition, a brand new set of powerful code similarity and recognition tools will be introduced as part of the new Knowledge Engine (see below).
Malware analysis helper tools: In 2026, we will bring a new set of optional companion tools designed to streamline malware analysis workflows and amplify the power of IDA.
New Reverse Engineering Tools
This year, reverse engineers - and their agentic servants - will be offered a new set of tools designed to unlock productivity while maintaining full explainability, deterministic outcomes, high processing speed, and low processing costs.
These tools, which we refer to as Engines, are not simple features or plugins, but standalone, analysis backends designed to power advanced reverse engineering workflows at scale.
They are multi-binary, multi-user, multi-threaded enterprise servers that:
- expose powerful, domain-driven, human- and LLM-friendly APIs
- are shipped with IDA plugins for a seamless experience within IDA, and can be connected to other RE platforms, such as Ghidra
- are fully containerized for easy on-premise or in a private cloud deployment; in a later stage, they will also be available in SaaS mode through the Hex-Rays community instances
- are extensible through Hex-Rays’ Engine SDKs that can be programmed in Python, Rust, C++, or with your favorite coding agents
- are highly scalable with respect to the number of binaries, users, and processes.
IDA Knowledge Engine
What if most of the code you needed to analyze was already neatly analyzed?
Think FLIRT & Lumina on steroids.
Think of an engine that can identify known and unknown code with accuracy and explain why matches occur. This is exactly what the IDA Knowledge Engine is designed to do.
The Knowledge Engine enables users to leverage their own collection of binaries, past analyses, and existing IDBs. At the same time, it comes with a large corpus of curated signatures, symbols, hashes, and vector embeddings drawn from known OS libraries, goodware, malware, and firmware.
To maximize accuracy and explainability, multiple similarity algorithms operate at different abstraction levels:
- Byte level (FLIRT & FLIRT 2.0)
- Disassembled function level (Lumina)
- Pseudo-code level (FLAKE, BBSH)
- Semantic level
Interested? Contact product@hex-rays.com to participate in our beta program.
We’re also providing in-person demos at RE//verse in Orlando, FL, 5 - 7 March 2026.
IDA Semantic Engine
Even with the help of agents, binary analysis can be slow and tedious.
If we had to guess, you’ve likely spent hours tracing data flows, writing fragile scripts that break on every new binary, or experiencing click-through hell, while navigating hundreds of cross-references.
What if you, and your favorite agents, could use a tool like CodeQL (which source code teams have had for years) to simply express sophisticated security queries and obtain results within seconds?
This is what the IDA Semantic Engine is all about. It enables users to unearth salient facts about binaries by formulating powerful, expressive queries, such as:
- “Show me inter-procedural data flows bidirectionally”
- “Find me implementations of cryptographic algorithms”
- “Identify functions that dereference a pointer returned by malloc without a NULL check”
- …
Some of us believe Binary Semantic Analysis is the biggest shift since a) decompilation and b) the use of AI.
Curious? Contact product@hex-rays.com to participate in the beta product trials.
We’re also available to give a sneak preview at RE//verse in Orlando, FL, 5 - 7 March 2026.
Quick Recap
If 2025 was about execution, 2026 is about expansion.
This year, we’re investing in deeper analysis, broader platform coverage, smarter automation, and infrastructure designed to support large-scale reverse engineering workflows.
The tools are evolving, the workflows are evolving, and so are we.
P.S. A birthday present to Madame de Maintenon…
IDA turns 35 this year and we feel that reverse engineers deserve a new generation of professional solutions to celebrate the milestone.
Stay tuned…
