Skip to main content
QWED v11.0.0 is the current release. It fixes ten advisories. The code, SQL, maths, image and guard checks admit only what they can show safe. QWEDLocal.verify, verify_math and verify_logic no longer report a claim as verified. /verify/logic proves a claim instead of reporting satisfiability. The Go module path ends in /v11. Read the 11.0.0 upgrade steps before you upgrade.

What is QWED?

QWED (Query With Evidence & Determinism) is a trust boundary for AI systems:
  • LLMs can translate user intent into structured claims.
  • QWED verifies those claims with deterministic engines before execution or response.
  • You get proof-backed outcomes instead of probability-only confidence.
QWED is designed for LLM verification, AI agent security, verified tool calls, prompt injection defense, and deterministic transaction verification in high-stakes workflows.
“Do not trust generated output. Verify it.”

Explore common verification problems

LLM verification

Learn how formal verification for LLMs differs from prompting, RAG, and output formatting.

AI agent security

Add pre-execution checks, policy enforcement, and budget controls for agent actions.

Prompt injection defense

Harden your stack against prompt injection, exfiltration, and unsafe execution paths.

QWED MCP

Give MCP clients access to QWED’s deterministic verification libraries.

When to use QWED first

Math and logic

Verify equations, constraints, and logical claims before they reach users.

Code and SQL

Catch unsafe patterns, injection risks, and structural errors before execution.

Agent tool calls

Inspect actions and payloads before external systems are touched.

High-stakes workflows

Add deterministic checkpoints for finance, legal, tax, and regulated systems.

Quick start (5 minutes)

The examples below use the qwed CLI that this installs. The qwedai/qwed-verification:11.0.0 Docker image runs the API server instead; see Self-hosting.
To prove a claim from its conditions with Z3, see Logic verification in the quick start: given x > 5, the claim x > 3 follows, and x > 10 comes back with a counterexample such as x = 6.
1

Install and configure

Follow Installation and set your provider with LLM configuration.
2

Run first verification

Use Quick start to validate math, logic, code, and SQL.
3

Integrate into production flow

Verification engines at a glance

Math

SymPy-based symbolic verification.

Logic

Z3 SAT/SMT verification with models.

Code

AST and symbolic checks for risky behavior.

SQL

Parser-backed SQL safety and validation.

Schema

Type and shape validation for structured outputs.

Taint

Data-flow tracking for untrusted inputs.

See all engines

Explore core, analysis, and specialized engines.

What’s new in v11

Admission only on proof

From v11.0.0, the code, SQL, maths, image and guard checks admit only what they can show safe, which fixes ten advisories. /verify/code never admits JavaScript, TypeScript, Java or Go. /verify/sql admits only reads of tables and known functions. /verify/logic decides whether a claim follows for every value. A claim that follows is VERIFIED, and a claim proven false is also VERIFIED, with is_valid: false and a counterexample. Every translation is screened for prompt injection in one place, before any provider is called. Read the 11.0.0 notes →
From v11.0.0, QWEDLocal.verify, verify_math and verify_logic return verified False for every claim, and qwed verify exits 1. Send a maths claim as an equation to /verify/math. A satisfiable set of constraints is no longer VERIFIED on /verify/logic. Some SQL and maths that 10.0.0 admitted is now refused. The Go module path moves to /v11. Read the upgrade steps →
  1. Core concepts
  2. Architecture overview
  3. SDKs overview
  4. API overview
  5. Integration guide