Summary

arimxyer/models offers a unified TUI and CLI for discovering, comparing, and tracking AI models, benchmarks, and coding agents. It targets technical users needing a consolidated view of the AI landscape to facilitate evaluation and adoption decisions, aggregating data from multiple sources into an accessible format.

How It Works

Built in Rust, the application provides interactive TUI and scriptable CLI access. It ingests data from models.dev (AI models), Artificial Analysis (benchmarks), and a curated catalog plus GitHub API (coding agents). The TUI features distinct tabs for Models (3,000+ models, capability indicators), Benchmarks (performance comparison via plots/tables), and Agents (coding assistant tracking). The CLI enables direct data access for automation.

Quick Start & Requirements

Installation options include cargo install modelsdev (Rust/Cargo required), Homebrew, Scoop, Arch AUR, native .deb/.rpm packages, or pre-built binaries from GitHub Releases. Source builds require cargo build --release. No specific hardware prerequisites like GPUs are listed for the tool itself.

Highlighted Details

• Models Tab: 3,000+ models, 85+ providers, capability indicators (Reasoning, Tools, etc.), provider categorization.
• Benchmarks Tab: ~400 entries from Artificial Analysis, auto-updating data (30 min refresh), comparison modes (tables, scatter, radar).
• Agents Tab: Tracks 12+ coding assistants with version detection, GitHub integration, and customizable lists.
• CLI: Supports scripting via JSON output for all model commands.
• Packaging: Native Linux packages (.deb, .rpm) for x86_64 and aarch64.

Maintenance & Community

Maintained by @Dominiquini, the project welcomes contributions via PRs under the Contributor Covenant Code of Conduct. Community channels like Discord/Slack are not explicitly linked, but Nix flake integration is noted as a potential roadmap item.

Licensing & Compatibility

Released under the permissive MIT license, allowing broad compatibility for commercial use and integration within closed-source applications.

Limitations & Caveats

README notes that videos/screenshots are out-of-sync with the current app state due to rapid development. Custom agent tracking requires manual TOML configuration. No other significant limitations are detailed.