Gortex
Gortex is open-source code intelligence that answers AI coding agent queries in one knowledge-graph call — up to 50x fewer tokens.
If your coding agent re-reads the same files on every edit loop, Gortex is the cheapest fix you'll find this quarter. The token math is real — 11,480 down to 688 on its own worked example — and it costs nothing to try because it's Apache 2.0. The honest caveat: on a small repo where grep plus two reads already answers everything, the graph is infrastructure you won't feel.
Verified 4d ago · liveness 72/100 · cite: rightaichoice.com/tools/gortex
- Developers running Claude Code, Cursor, or Copilot on large codebases where repeated file reads burn context
- Teams on monorepos with heavy cross-file and cross-repo dependencies
- Engineers building custom MCP-based coding tools who need a programmatic graph API
- Open-source maintainers who want local, zero-dependency code intelligence with no cloud component
- Anyone not using an AI coding agent — the graph is served to agents, not to you directly
- Teams that require a cloud-hosted or SaaS deployment of the index
- Projects in languages outside the 257 supported set without supplying a custom grammar
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Gortex if you are not using an AI coding agent, if you need a cloud-hosted code intelligence service, or if your codebase is small enough that grep plus two file reads already answers the question.
Plain `gortex install` wires MCP config, skills, and slash commands into your user-level agent directories — budget the time to review those generated files before your team trusts them.
Gortex is free and open source under Apache 2.0, which puts it below every paid code-intelligence peer — Sourcegraph and similar hosted platforms charge per seat or per usage, and commercial AI coding agents bill per token. Its real cost is local compute and one setup afternoon. For a monorepo team already paying per token, that trade usually pays back in the first week of agent use.
In short
Gortex — Gortex is open-source code intelligence that answers AI coding agent queries in one knowledge-graph call — up to 50x fewer tokens. Best for Developers running Claude Code, Cursor, or Copilot on large codebases where repeated file reads burn context, Teams on monorepos with heavy cross-file and cross-repo dependencies, Engineers building custom MCP-based coding tools who need a programmatic graph API. Free to use.
What's new in Gortex
Checked 3 days agoAcross the latest 3 updates: 2 feature updates and 1 changelog entry.
MCP 2026 Streamable HTTP transport supported on the server and daemon
Both `gortex server` and the daemon's `--http-addr` bind expose /mcp — the Streamable HTTP transport the June 2026 MCP spec locks in — handling POST/GET/DELETE, per-request state via Mcp-Session-Id, JSON-RPC batching, and server-initiated SSE notifications.
v0.62.0 — 21-tool MCP facade and GCX1 wire format
Named agents get a compact 21-tool facade (facade-v1) with the remaining tools behind tools_search, and the round-trippable GCX1 text format cuts a median 27.4% of tokens versus JSON.
Sigstore-signed releases with SLSA Level 3 provenance
Releases ship as signed native binaries for macOS, Linux, and Windows on amd64 and arm64, with the installer verifying SHA256 against checksums.txt and cosign when available.
What people actually say about Gortex — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
58 mentions across 5 sources (Hacker News, YouTube, Bluesky, GitHub, Lemmy) · researched Jul 15, 2026.
Average across the 5 sources that answered — each source counts once, not each post.
- +Up to 50x token reduction per agent edit via graph queries.
- +100+ MCP tools out of the box for deep code analysis.
- +Supports 257 languages with tree-sitter and regex parsers.
- +Runs locally, one binary, zero external dependencies.
- +Live buffer overlays for unsaved changes via shadow graphs.
- −Consumes 12GB+ per worktree — OOM on medium repos.
- −Codex hooks skip 91% of calls and PostToolUse HTTP fails.
- −Hooks cause agents to run invalid CLI commands.
- −No cross-repo contract recognition or business flow support.
- −Small community — limited support and bug fix velocity.
- • Optional semantic search requires running local models (MiniLM/Ollama/OpenAI) which need extra compute and memory.
Viability Score
How well maintained and how widely used is Gortex? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Knowledge-graph indexing of files, symbols, imports, calls, modules, SQL tables, and K8s resources
- 257-language parsing across three tiers: ~30 tree-sitter grammars, ~60 regex, ~165 signature-only
- Semantic search default-on: hybrid BM25 + vector with RRF fusion and a baked 3.8 MB GloVe-50d table
- Optional deeper embeddings via MiniLM, Ollama, or OpenAI
- Live editor-buffer overlays through shadow-graph sessions for unsaved files
- Speculative edit previews with per-step diagnostics delta and blast-radius rollup
- Dataflow and taint tracing for security audits without loading a sandbox
- MinHash clone detection with LSH banding and Jaccard similarity edges
- Structural unsafe-pattern scanning via tree-sitter S-expressions with seven detectors
- Concurrency analysis for race writes, unclosed channels, and a 0-100 health score graded A-F
- Graph-grounded PR review producing BLOCK/REVIEW/APPROVE verdicts with line-anchored comments
- 17-server LSP bridge for diagnostics and atomic code actions
- 21-tool MCP facade fronting a 175-tool catalogue with instruction profiles
- MCP 2026 Streamable HTTP transport with per-request Mcp-Session-Id state
- GCX1 compact text format, median 27.4% fewer tokens than JSON
About Gortex
Gortex is an Apache 2.0, local-first code intelligence layer for AI coding agents. It indexes your repository into an in-memory knowledge graph and serves it to agents over MCP, HTTP, and a web UI, so a question that would normally cost half a dozen file reads plus a grep resolves in a single graph query — with callers, callees, imports, blast radius, and interface implementors attached. Its own worked example drops a "make an edit" loop from 11,480 tokens to 688, a 94% reduction. The current v0.62.0 build ships as one static binary with zero external services on macOS, Linux, and Windows (amd64 and arm64). It parses 257 languages through a three-tier extractor: about 30 bespoke tree-sitter grammars, roughly 60 regex for niche or legacy code, and around 165 forest-backed signature-only parsers, plus Jupyter and Databricks notebooks indexed cell-by-cell. Releases are Sigstore-signed with SLSA Level 3 build provenance, and the installer verifies SHA256 against checksums.txt. On top of the graph it layers semantic search (hybrid BM25 plus vector with a baked 3.8 MB GloVe-50d table, with MiniLM, Ollama, or OpenAI opt-in), shadow-graph overlays for unsaved editor buffers, speculative edit previews that never touch disk, dataflow and taint tracing, MinHash clone detection, structural unsafe-pattern scanning, race-write and unclosed-channel analyzers, a 0-100 health score, graph-grounded PR review, and a 17-server LSP bridge. A 21-tool MCP facade fronts a 175-tool catalogue, and the GCX1 compact wire format cuts a median 27.4% of tokens versus JSON. It is wired for named agents including Claude Code, Cursor, GitHub Copilot, Aider, Continue.dev, Windsurf, and Zed, with 19 agents supported in total. Everything runs in-process on your machine with no cloud component. For privacy-sensitive teams, large monorepos, and anyone watching token spend, that is the whole pitch: your agent gets relational answers instead of a stack of file reads, and your source never leaves the
Behind the Verdict
We'd reach for Gortex the moment an agent's context bill or latency starts tracking repo size rather than task difficulty. That's the signal. On a monorepo with deep cross-file dependencies, the difference between "Read internal/watcher/watch.go (412 lines)" five times over and one explore call that returns signatures, callers, and an interface match is not marginal — it's the difference between an edit loop that fits in context and one that thrashes. Pick it when you're running Claude Code, Cursor, or Copilot on a large, dependency-heavy tree, or when a security review needs taint tracing and unsafe-pattern scans without shipping source to a sandbox. The dataflow, clone-detection, and race-write analyzers are the parts that surprise people — they're graph work, not LLM work, so they're deterministic and cheap. Where it bites: language coverage is tiered, and the ~165 signature-only parsers won't give you the deep resolution the ~30 tree-sitter grammars do. If your stack lives in that long tail, expect thin results until you drop a custom grammar in via .gortex.yaml. Pass on it for tiny repositories. These tools have a coverage floor — every new tool added to an agent is a routing decision the model can get wrong — and if two file reads answer your question, the graph is ceremony. We'd also flag the LSP and cross-repo machinery as real value only if you actually have orphans and mismatches to catch across a multi-service tree; single-service codebases won't exercise it. Closest alternative depends on what you're optimizing. Aider's repo-map solves the same token-pressure problem with a flatter representation and no MCP surface — simpler, less relational. Gortex is the better fit when your questions are about structure (who calls this, what breaks if I change it,
Researching Gortex? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Gortex actually fits — and what changes day-one when you adopt it.
You ask the agent to change a watcher callback. With Gortex wired via `gortex mcp --watch`, the agent runs one explore query instead of six Reads and a Grep.
Outcome: The edit loop lands at 688 tokens instead of 11,480 (−94%), with the satisfied interface and a 4-symbol, 2-test blast radius attached before any file is touched.
You run trace(taint) for source-to-sink sweeps and analyze(unsafe_patterns) to fan seven detectors — SQL string concatenation, weak crypto, hardcoded secrets — over the whole file set in one call.
Outcome: You get ranked taint paths and unsafe-pattern hits without provisioning a sandbox or uploading source anywhere.
With the graph built across repos, cross-repo contract detection surfaces HTTP, gRPC, GraphQL, pub/sub, and Temporal connections, and the review returns a BLOCK/REVIEW/APPROVE verdict grounded in the graph.
Outcome: Callers, callees, and interface implementors are enumerated up front, so review catches the missed implementor instead of the build.
Use Cases
- Cut token consumption on agent edits in a large repo — Gortex's example drops an edit loop from 11,480 to 688 tokens
- Answer "who calls this function and what interfaces does it satisfy" in one graph query with callers and blast radius attached
- Search your whole repository semantically with hybrid BM25 + vector ranking
- Preview an edit's impact — diagnostics delta and cumulative blast radius — before it touches disk
- Detect clone duplicates, unsafe patterns, race writes, and unclosed channels across the codebase
- Index Jupyter and Databricks notebooks cell-by-cell for data science workflows
- Review a PR with a graph-grounded BLOCK/REVIEW/APPROVE verdict
- Serve one shared graph to every editor window through the daemon
Models Under the Hood
as of 2026-10-17
Limitations
- Gortex is a local, in-memory code intelligence tool for repositories — it is aimed at people already driving an AI coding agent, since its token savings come from replacing the agent's file reads with graph queries.
- It ships as one static binary per machine (macOS, Linux, Windows) with a per-repo setup before savings show up.
- Language coverage is three-tier: ~30 languages get bespoke tree-sitter extraction, ~60 are regex-based, and ~165 are forest-backed signature-only, so precision on exotic languages is shallower than on core ones.
- Its MCP surface is a 21-tool facade fronting a larger 175-tool catalogue, which must be searched to reach the rest.
as of 2026-10-05
Verification history
We have re-verified Gortex 9 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 9 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Gortex tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Open Source (Apache 2.0)
$0
Where the pricing makes sense
The company stage and team size where Gortex's pricing actually pencils out — and where peers do it cheaper.
Gortex is free and open source under Apache 2.0, which puts it below every paid code-intelligence peer — Sourcegraph and similar hosted platforms charge per seat or per usage, and commercial AI coding agents bill per token. Its real cost is local compute and one setup afternoon. For a monorepo team already paying per token, that trade usually pays back in the first week of agent use.
Setup time & first value
How long it actually takes to get something useful out of Gortex — broken out by persona, not the marketing-page minute.
Once per machine: one command (`curl -fsSL https://get.gortex.dev | sh`, Homebrew, Scoop, apt/dnf/apk, or `go install`), then `gortex install` for user-level MCP wiring. Once per repo: `gortex init` — about 15 minutes end to end per the vendor's walkthrough. Add `--analyze` if you want an indexed CLAUDE.md. First useful graph query typically lands inside that same session.
Switching to or from Gortex
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From plain agent file reads: install Gortex, run `gortex init`, and the agent switches from Read/Grep loops to explore/search calls.
- →From Aider's repo-map: keep Aider and point it at the Gortex MCP surface so relational queries replace the flat map.
- →From repo-local scripts: `gortex init --hooks-only` installs repo-local hooks without regenerating community skills.
- →From a single-editor setup: switch `gortex mcp` to `gortex daemon` to share one graph across every editor window.
- ↗To Aider's built-in repo-map: drop the MCP wiring and let Aider manage context from its own map.
- ↗To a hosted code intelligence platform: port your symbol and dependency queries to the vendor's API; Gortex's graph is local-only.
- ↗To raw grep/ripgrep workflows: remove the MCP server config and the per-repo .mcp.json wiring; no data migration is needed.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “Gortex”, and we withheld 6: 6 could not be judged, because “Gortex” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Gortex.
Official links
Tools that pair well with Gortex
Common stack mates teams adopt alongside Gortex, with the specific reason each pairing earns its keep.
Continue
Open-source AI coding agent for VS Code and JetBrains — free to fork since Cursor acquired Continue in January 2026.
Sourcegraph Cody
Sourcegraph Cody: an AI coding assistant that grounds chat, completions, and agent workflows in your entire multi-repo codebase.
Refact.ai
Open-source autonomous AI coding agent that plans, executes, and deploys tasks inside VS Code and JetBrains IDEs.
Featured Head-to-Head Comparisons
Gortex vs Spider Cloud
These tools solve opposite problems: Gortex slashes token costs for AI coding agents by indexing your local codebase into a knowledge graph, while Spider Cloud feeds AI agents live web data via a fast scraping API. Pick Gortex if you're a developer wrestling with large repositories and high token bills; pick Spider Cloud if your AI needs real-time web content for RAG or research.
Gortex vs Voyage Ai
Voyage AI and Gortex target completely different problems: Voyage AI is for enterprise retrieval pipelines needing high-accuracy embeddings and rerankers on specialized domains, while Gortex slashes token costs for developers using AI coding agents by building an local knowledge graph. If you need to improve search accuracy on dense legal/financial documents, choose Voyage AI. If you want your AI coding assistant to understand your entire codebase without burning tokens, Gortex is the clear winner — and it's free.
Gortex vs Temporal Ai
Gortex and Temporal AI solve fundamentally different problems. Pick Gortex if your primary pain point is token cost and latency when using AI coding agents on large codebases — it's free, local, and plugs directly into your existing agent tools. Choose Temporal AI if you need a battle-tested durable execution platform for building reliable, long-running AI agent workflows that must survive crashes and human-in-the-loop pauses; it's more complex but enterprise-ready. For most AI-assisted developers, Gortex is a quick win; for teams building production AI agents, Temporal is the backbone.
Gortex vs Poolside Ai
If you're an individual developer or team using AI coding agents and want to slash token costs and latency by replacing file reads with graph queries, Gortex is the free, immediate-win choice. For enterprises in regulated industries needing custom, open-weight models with multi-agent orchestration, sandboxed execution, and auditability—deployable in air-gapped environments—Poolside AI's Laguna models and platform are purpose-built. Choose based on whether your priority is cost-efficient local code intelligence (Gortex) or governed, long-horizon agentic coding at scale (Poolside).
Alternatives to Gortex
View allContinue
Open-source AI coding agent for VS Code and JetBrains — free to fork since Cursor acquired Continue in January 2026.
Sourcegraph Cody
Sourcegraph Cody: an AI coding assistant that grounds chat, completions, and agent workflows in your entire multi-repo codebase.
Frequently Asked Questions
Best-of guides
Used Gortex? Help shape our editorial sentiment research.