{"slug": "litespeed-a-local-coding-agent-with-multi-model-workflows", "title": "Litespeed – a local coding agent with multi-model workflows", "summary": "BerriAI released Litespeed, a local coding agent for the terminal and browser built around multi-model workflows, distributed as a macOS package with runtimes included so no Node or npm setup is required. Litespeed connects through LiteLLM, OpenAI-compatible APIs, native Anthropic, or ChatGPT device sign-in, and its recommended Sidekick Fusion arrangement pairs a strong driver model for planning and review with a cheaper sidekick for implementation, while the experimental LiteFusion routes 63 task categories to specific models and reasoning levels. The agent runs a local backend automatically, serves a browser UI at http://localhost:3210, and keeps saved sessions and settings outside the application package.", "body_md": "A local coding agent for your terminal and browser, built around **multi-model workflows**. Use one model or combine a driver with sidekicks, workers, or experts. Follow their work, approve changes, and review the result in one conversation. Pairing models lets a faster, cheaper model handle routine work while a stronger model handles planning or difficult tasks, which can reduce cost and wait time.\n\nConnect through LiteLLM, OpenAI-compatible APIs, native Anthropic, or ChatGPT device sign-in. Your project stays on your machine; prompts and selected context go to your chosen provider.\n\nDownload the **macOS package** with its runtimes included—no Node or npm setup:\n\n```\ncurl -fsSL https://github.com/BerriAI/litespeed/releases/latest/download/install.sh | sh\n```\n\nOpen a **new terminal**, then run it from **the project you want to work on**:\n\n```\ncd /path/to/your/project\nlitespeed\n```\n\nThe installer sets up the `litespeed` command for new terminal windows (zsh or Bash). If you use the same terminal window where you installed it, run `export PATH=\"$HOME/.local/bin:$PATH\"` first.\n\nLitespeed starts its local backend automatically. Updates appear in both UIs, or run `litespeed update`. Saved sessions and settings stay outside the application package.\n\nSee [installation, updates, and the source-build alternative](https://github.com/BerriAI/litespeed/blob/main/docs/installing.md). Packages support Apple silicon and Intel Macs. Git and your project's development tools remain separate.\n\nLitespeed has its own command, separate from the [LiteLLM gateway CLI](https://docs.litellm.ai/docs/proxy/management_cli). If you used an earlier version of this agent, follow the [upgrade guide](https://github.com/BerriAI/litespeed/blob/main/docs/upgrading.md) to carry over saved sessions and configuration.\n\n1. On your first launch, choose your setup, connect your **LiteLLM gateway base URL** and**API key** , and pick your models.**Sidekick Fusion is recommended:** choose a strong driver to plan and review, and a cheaper sidekick to handle implementation.**LiteFusion is experimental** ; its task-specific routing policy is available for evaluation. Choose**Single model** if you prefer one model for everything. Litespeed remembers the connection and models; running`litespeed` in another project opens chat directly. Gateways that do not require a key can leave it blank.\n2. Use **Models** or`/models` to change your arrangement.**Ask first** is the default;**Allow all tools** is available in permissions. The full setup is available from the web sidebar or`/setup` . If you already use Claude Code or Codex skills, /skills (or**Settings → Project profiles** ) offers**Import a Claude/Codex skill…** to copy one into the project — see[Project profiles and skills](https://github.com/BerriAI/litespeed/blob/main/docs/profiles.md) .\n3. Type a task. **Build** can edit files and run commands;**Plan** uses read-only tools. Type`/` for command suggestions in either client; use**↑/↓** ,**Tab** or**Enter** to complete, and**Esc** to dismiss.**Ctrl+P** opens terminal commands and navigation.\n\nSee the [terminal guide](https://github.com/BerriAI/litespeed/blob/main/docs/tui.md) for shortcuts, resuming sessions, and configuration.\n\nOnce Litespeed is running, open **[http://localhost:3210](http://localhost:3210)**. The browser and terminal share saved sessions, providers, and model settings.\n\nFor a web-only session, run `litespeed serve`. For development, use `npm run dev` from the checkout. See [development and updating](https://github.com/BerriAI/litespeed/blob/main/docs/development.md).\n\n| Architecture | How it works | \n|---|---|\n| **Single model** | One model investigates, implements, and checks the task. | \n| **Sidekick Fusion** · Recommended | A strong driver plans and reviews; a cheaper sidekick keeps context across handoffs. | \n| **LiteFusion** · Experimental | One persistent lead routes 63 task categories to specific models and reasoning levels, with one shared hard/escalation map. | \n| **Team Fusion** | A strong driver assigns fresh cheaper workers, runs independent work in parallel, and verifies the combined result. | \n| **Expert Fusion** | A cheaper driver coordinates fresh strong experts and verifies their work. Independent assignments can run in parallel. | \n\nChoose any connected model for each role. An optional **Planner model** handles Plan mode separately. Litespeed uses your last chosen model arrangement for new sessions across workspaces; existing sessions keep their settings. Cost and quality depend on the models and task. See [architecture details and limits](https://github.com/BerriAI/litespeed/blob/main/docs/architectures.md).\n\n**Shunt** sends large reads and routine generation to a separate model. Off by default; enable it in **Advanced settings** during setup or in Models, in either client. [How it works and measured results](https://github.com/BerriAI/litespeed/blob/main/docs/shunt.md).\n\n**Session goals** (`/goal`) keep an objective moving across turns. There is no turn limit unless you set one. [Context compaction](https://github.com/BerriAI/litespeed/blob/main/docs/context-management.md) makes room during long tasks automatically.\n\n- [Using Litespeed](https://github.com/BerriAI/litespeed/blob/main/docs/usage.md) : queue follow-ups, steer a response, answer questions, manage context, and customize projects.\n- [Providers](https://github.com/BerriAI/litespeed/blob/main/docs/providers.md) ·[CLI and scripting](https://github.com/BerriAI/litespeed/blob/main/docs/cli.md) ·[Terminal controls](https://github.com/BerriAI/litespeed/blob/main/docs/tui.md)\n- [Permissions](https://github.com/BerriAI/litespeed/blob/main/docs/permissions.md) ·[Undo/redo, recovery, and local data](https://github.com/BerriAI/litespeed/blob/main/docs/local-data.md)\n- [Project profiles and skills](https://github.com/BerriAI/litespeed/blob/main/docs/profiles.md) ·[MCP connections](https://github.com/BerriAI/litespeed/blob/main/docs/mcp.md) ·[Hooks and plugins](https://github.com/BerriAI/litespeed/blob/main/docs/design-hooks-plugins.md)\n- [Context management and harness comparison](https://github.com/BerriAI/litespeed/blob/main/docs/context-management.md) ·[Concurrent tasks](https://github.com/BerriAI/litespeed/blob/main/docs/concurrency.md)\n- [Agent memory](https://github.com/BerriAI/litespeed/blob/main/docs/memory.md) ·[History search](https://github.com/BerriAI/litespeed/blob/main/docs/search.md) ·[Research tasks](https://github.com/BerriAI/litespeed/blob/main/docs/delegation.md)\n- [Feature coverage and known gaps](https://github.com/BerriAI/litespeed/blob/main/docs/coverage.md) ·[Development and tests](https://github.com/BerriAI/litespeed/blob/main/docs/development.md)\n\nApproved commands run with your local user’s capabilities; permissions are not a sandbox. Keep the server local. Provider keys stay server-side, and provider usage may incur charges.\n\nLitespeed is licensed under [Apache-2.0](https://github.com/BerriAI/litespeed/blob/main/LICENSE). Bundled themes and fonts retain their original licenses; see [third-party notices](https://github.com/BerriAI/litespeed/blob/main/THIRD_PARTY_NOTICES.md).", "url": "https://wpnews.pro/news/litespeed-a-local-coding-agent-with-multi-model-workflows", "canonical_source": "https://github.com/BerriAI/litespeed/tree/main", "published_at": "2026-09-17 12:09:28+00:00", "updated_at": "2026-09-17 12:26:16.043310+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "developer-tools", "ai-products"], "entities": ["BerriAI", "Litespeed", "LiteLLM", "OpenAI", "Anthropic", "ChatGPT", "Sidekick Fusion", "LiteFusion"], "alternates": {"html": "https://wpnews.pro/news/litespeed-a-local-coding-agent-with-multi-model-workflows", "markdown": "https://wpnews.pro/news/litespeed-a-local-coding-agent-with-multi-model-workflows.md", "text": "https://wpnews.pro/news/litespeed-a-local-coding-agent-with-multi-model-workflows.txt", "jsonld": "https://wpnews.pro/news/litespeed-a-local-coding-agent-with-multi-model-workflows.jsonld"}}