OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Chronos — Kodezi

Chronos is Kodezi's proprietary language model for repository-scale code understanding and debugging. It is presented as a tool for diagnosing and fixing bugs across codebases, using adaptive graph-guided retrieval and persistent debug memory. The public GitHub repository contains research, benchmark tooling and evaluation results, but not the model.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Chronos — Kodezi — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Consider Chronos when repository-wide debugging is the task and you are assessing the model's reported results or pursuing access through Kodezi OS. Use the public repository's benchmark tools when the goal is to evaluate a different model; the Chronos model itself is not included.

Who it is for

  • Software teams interested in debugging assistance across multi-file codebases.
  • Developers and researchers evaluating code-debugging models with the published MRR benchmark tools.

Documented uses

  • Investigate bugs that require understanding code across a repository.
  • Evaluate a supplied model on the repository's MRR benchmark scenarios and analyze its results.
  • Review Chronos's published research materials and reported evaluation results.
Before you choose

Requirements & limits

The model is described as proprietary and available through Kodezi OS; the repository links to a waitlist at chronos.so. The model is not included in the public repository.

  • The research repository and the Chronos model are distinct: benchmark code and results are public, while the model is proprietary.
  • The documented MRR workflow requires Python 3.8 or later, Git, and a model identifier; it evaluates the user's model rather than providing Chronos.
  • Kodezi reports 80.33% on SWE-bench Lite and 67.3% debug success in its MRR materials; these are project-reported results.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run and analyze the MRR benchmark for another model

Use the repository's benchmark runner and analysis script to evaluate a model you provide.

Before you start

  • Python 3.8 or later
  • Git
  • A model identifier accepted by the benchmark runner

Steps

  1. Clone the chronos-research repository and change into its directory.

  2. Install dependencies with pip install -r requirements.txt.

  3. Run benchmarks/run_mrr_benchmark_2025.py with --model and --scenarios; the documentation describes 100 scenarios as a subset and 500 as the available sample evaluation.

  4. Pass the resulting directory to benchmarks/analyze_results.py using --results_dir.

Expected outcome

Benchmark results are produced for the supplied model and can be analyzed with the repository's results tool.

Guide limitations

  • Chronos is proprietary and is not included in the repository; this workflow evaluates a model supplied by the user rather than running Chronos.
  • The documented sample evaluation contains 500 scenarios.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
Kodezi/Chronos · Official homepage title · retained in its source language
How we review sourcesSuggest a correction