# Alternatives to WTAgent > Source: https://theagentsindex.com/wtagent/alternatives (current datasheets, evidence-ranked) Candidates share the baseline category or are previously researched peers. Requirements, preserved strengths, explicit trade-offs and research coverage decide the order; paid tier does not. ## Shortlist ### 1. Cline Open-source coding agent for VS Code, JetBrains, the terminal and CI, running on whichever model you bring. **Research coverage:** 34/35 criteria **Where it improves on WTAgent:** - [Runs autonomously: yes — Auto-approval and headless mode let a task run start to finish with nobody watching, including scheduled cron runs and CI jobs; the vendor warns to use a clean branch.](https://docs.cline.bot/usage/cli-overview) - [Agent permissions: yes — Approval is per tool call across eight named categories - reading, editing, safe commands, all commands, browser, MCP - each toggled separately, with outside-workspace access gated behind the base toggle.](https://docs.cline.bot/features/auto-approve) - [Open models: yes — Open-weight models are a first-class path: a curated set ships inside the ClinePass subscription, and Ollama, LM Studio or any OpenAI-compatible endpoint runs weights you host.](https://cline.bot/cline-pass) **Trade-offs against WTAgent:** - [Getting out: partial — Task history is plain JSON on your own disk and can be mirrored to your S3 or R2 bucket, but the terms make fees non-refundable and give nothing back for unused credits on cancellation.](https://cline.bot/tos) [View Cline](https://theagentsindex.com/cline) ### 2. goose Open-source local AI agent — desktop app, CLI and ACP server — running on the LLM provider you bring. **Research coverage:** 34/35 criteria **Where it improves on WTAgent:** - [Runs autonomously: yes — Autonomous mode is the shipped default and needs no human in the loop; goose run takes an instruction file, allows 1000 turns without user input, and recipes can be scheduled by cron.](https://goose-docs.ai/docs/guides/managing-tools/goose-permissions) - [Agent permissions: yes — Four permission modes span full autonomy to chat-only, with per-tool Always allow / Ask before / Never allow rules, switchable mid-session; the default grants shell and file writes without asking.](https://goose-docs.ai/docs/mcp/developer-mcp) - [Open models: yes — Open-weight models run locally through Ollama, LM Studio, Docker Model Runner, Ramalama or goose's own local-models command, and remotely through gateways such as Novita, OVHcloud and Cerebras.](https://goose-docs.ai/docs/getting-started/providers) **Trade-offs against WTAgent:** - [How it meters: partial — goose itself charges nothing, so the running cost is entirely your LLM provider's metered token or credit spend; no calculator or cost estimate is published and the vendor only tells you to keep provider credits topped up.](https://goose-docs.ai/docs/getting-started/installation) [View goose](https://theagentsindex.com/goose) ### 3. Kilo Code Open source AI coding agent for VS Code, JetBrains, CLI and cloud, with 500+ models at provider rates. **Research coverage:** 35/35 criteria **Where it improves on WTAgent:** - [Runs autonomously: yes — Cloud Agents run long tasks unattended in isolated containers and open PRs.](https://kilo.ai/) - [Open models: yes — Open-weight hosted models plus fully local models through Ollama or LM Studio.](https://kilo.ai/) - [In your editor: yes — VS Code and every VS Code-based editor, plus all JetBrains IDEs.](https://kilo.ai/) **Trade-offs against WTAgent:** - [Self-hosted: partial — The MIT clients can be forked and self-hosted and inference can be fully local, but Gateway, Cloud Agents and team administration are Kilo-hosted.](https://kilo.ai/) - [Getting out: partial — Usage data has documented export endpoints and credit purchases can be refunded (the 5% fee is not); the site states nothing about credits expiring.](https://kilo.ai/pricing) [View Kilo Code](https://theagentsindex.com/kilo-code) ### 4. Letta Stateful agent harness, cloud and SDK for agents with persistent, git-versioned memory. **Research coverage:** 35/35 criteria **Where it improves on WTAgent:** - [Runs autonomously: yes — Headless CLI runs, scheduled prompts and subagents execute without a human turn; unrestricted mode auto-allows tools.](https://docs.letta.com/configuration/subagents/index.md) - [Agent permissions: yes — Four modes from strict to unrestricted, plus allow/deny tool patterns and a cross-agent memory guard.](https://docs.letta.com/configuration/permissions/index.md) - [Open models: yes — Ollama, LM Studio and llama.cpp endpoints are local-only integrations; DeepSeek, Mistral and Together are also listed.](https://docs.letta.com/configuration/models/index.md) **Trade-offs against WTAgent:** - [Audit log: no — No administrative audit log is published; the Enterprise list stops at RBAC, SSO, quotas and dedicated support.](https://docs.letta.com/pricing/index.md) - [Getting out: partial — Memory is git-backed files you can inspect and sync to your own GitHub repo; credit expiry is never stated.](https://docs.letta.com/handbook/index.md) [View Letta](https://theagentsindex.com/letta) ### 5. OpenHands Open-source software engineering agent with GUI, CLI, SDK and an enterprise control plane. **Research coverage:** 34/35 criteria **Where it improves on WTAgent:** - [Runs autonomously: yes — Event-triggered runs complete the task with nobody in the loop, and headless CLI mode forcibly approves every action.](https://docs.openhands.dev/openhands/usage/cli/headless.md) - [Agent permissions: yes — Three confirmation policies ship in the SDK, from approving every action to only risky ones judged by a security analyzer.](https://docs.openhands.dev/sdk/guides/security.md) - [Open models: yes — Open-weight models are supported first-class: a recommended local model plus documented LM Studio, Ollama, vLLM and SGLang serving paths.](https://docs.openhands.dev/openhands/usage/llms/local-llms.md) **Trade-offs against WTAgent:** - [Open source: partial — The core is MIT-licensed open source, but the Enterprise control plane is described only as source-available under a commercial licence.](https://www.openhands.dev/pricing) - [Getting out: partial — You can manage and delete stored content, and deletion is by email request; no bulk data export or credit-expiry terms are published.](https://www.openhands.dev/privacy) [View OpenHands](https://theagentsindex.com/openhands) ### 6. Aider Open-source terminal AI pair programmer that edits files in your local git repo and commits the changes. **Research coverage:** 34/35 criteria **Where it improves on WTAgent:** - [Open models: yes — Local open-weight models are explicitly supported through Ollama, LM Studio and any OpenAI-compatible endpoint, so inference can stay on your own hardware.](https://aider.chat/docs/llms.html) - [In your editor: partial — There is no editor plugin; instead a file watcher picks up AI comments you type in whatever IDE you use, so the agent itself still runs as a separate terminal process.](https://aider.chat/docs/usage/watch.html) - [In your pipeline: partial — Non-interactive runs are scriptable from a shell loop with --message and --yes, but nothing on the site documents acting on a pull request or a hosted CI integration.](https://aider.chat/docs/scripting.html) **Trade-offs against WTAgent:** - [Ships it: no — The documented end of the loop is a git commit in your local repo; no page on the site describes deploying, pushing a release or publishing to an environment.](https://aider.chat/) - [How it meters: partial — The tool itself costs nothing; what you pay is token consumption at whichever LLM provider your key belongs to, and the docs treat context size as a direct cost to you.](https://aider.chat/docs/faq.html) [View Aider](https://theagentsindex.com/aider) ## Method Unknown data is never treated as a benefit or a trade-off. Featured placement is labelled but cannot alter recommendation order.