Chatluna
Open-source Koishi plugin that adds multi-model LLM chat, agent tools and text/image/voice replies to your bot
If you already run Koishi, ChatLuna is the plugin to reach for. An agent sandbox with MCP servers, web search, page browsing, file read/write and scheduled tasks is rare in a chat plugin, and the renderer pipeline turns model replies into text, images or voice while per-group presets give each room its own prompt and world book. The trade-off is structural: ChatLuna requires a Koishi installation, and cost and rate limits come from the model platforms whose keys you supply. Not on Koishi? Open WebUI or Jan stand alone but give up ChatLuna's tool sandbox and Koishi-native group management.
Verified 2d ago · liveness 66/100 · cite: rightaichoice.com/tools/chatluna
- Koishi bot developers building multi-model community chatbots
- Self-hosters who want one extensible LLM layer instead of many single-provider plugins
- Developers needing agentic tooling with a sandbox, MCP and shell access
- Group and community operators who want per-channel presets and API call limits
- Anyone not running the Koishi framework — ChatLuna installs as a Koishi plugin, not a standalone app
- Teams needing a hosted desktop or mobile chat application
- People looking for built-in fine-tuning, model training or model hosting
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip ChatLuna if you don't run a Koishi bot — it installs as a Koishi plugin, so there is no standalone app, no hosted signup, and model access depends on API keys you supply yourself.
Your real bill comes from the model platforms whose API keys you plug in, not from ChatLuna, so a busy group with an agent loop can run up spend on DeepSeek, OpenAI or Gemini that never appears in the plugin itself.
ChatLuna itself is released under CC-BY-SA-4.0 at no cost, so the economic comparison is against other Koishi chat plugins (also free, but usually single-provider) and against standalone clients like Open WebUI or Jan — those are free too, but you trade away ChatLuna's Koishi-native group presets and agent tool sandbox. Your actual spend is set by whichever model platform's API keys you plug in.
In short
Chatluna — Open-source Koishi plugin that adds multi-model LLM chat, agent tools and text/image/voice replies to your bot. Best for Koishi bot developers building multi-model community chatbots, Self-hosters who want one extensible LLM layer instead of many single-provider plugins, Developers needing agentic tooling with a sandbox, MCP and shell access. Free to use.
What's new in Chatluna
Checked 2 days agoAcross the latest 2 updates: 2 changelog entries.
Docs refreshed for agent capabilities and wording
The documentation was updated to align its agent-capability descriptions with the shipped tool set and tidy the wording across the guide.
Documentation site initialised on VitePress with full user guides
ChatLuna's docs moved to a VitePress site with structured user guides covering setup, model platforms, agent tools and presets; the sidebar was restructured in a later pass.
What people actually say about Chatluna — is it worth it?
We scanned public community sources for Chatluna on Jul 28, 2026 and could not establish that the discussion we found is about this tool rather than something else sharing its name. Our own analysis of that scan says the posts were off-subject. Rather than publish a sentiment score built on the wrong subject, we publish nothing here and re-run the scan.
Viability Score
How well maintained and how widely used is Chatluna? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Multi-model integration: DeepSeek, OpenAI, Google Gemini, Anthropic Claude and more
- Additional adapters for Qwen, Dify, ChatGLM, iFlytek Spark, ERNIE, Hunyuan, Moonshot, Ollama, RWKV
- Built on LangChain with a reusable model-service API for Koishi plugin developers
- Agent mode with sandboxed tool execution, enabled by installing additional plugins
- MCP (Model Context Protocol) support for external tool servers
- Web search tool for real-time queries inside chat
- Page browsing and arbitrary URL request tools
- OpenAPI request tool for hitting documented endpoints
- File read and write tools inside the agent sandbox
- Scheduled task tool for timed agent runs
- Koishi shell command execution tool
- Text-to-image tool and simple audio generation tool
- Deliberate-thinking and TODO tools for multi-step agent tasks
- Group moderation (mute) tool and active memory tool
- Per-group/per-room presets with render templates and world books
About Chatluna
ChatLuna is an open-source Koishi plugin that gives your bot a single LLM chat layer instead of a separate plugin per provider. Install it in Koishi, supply the API keys you already pay for, and you can route DeepSeek, OpenAI, Google Gemini, Anthropic Claude, Qwen, Dify, ChatGLM, iFlytek Spark, ERNIE, Hunyuan, Moonshot, Ollama and RWKV through one configuration. It is built on LangChain and exposes a reusable model-service API, so long-term memory, knowledge-base and renderer plugins can build on top of the same layer. The differentiator is Agent mode: install the additional agent plugins and the model gets a sandbox plus a tool set covering MCP servers, web search, page browsing, URL and OpenAPI requests, file read/write, scheduled tasks, Koishi shell command execution, text-to-image, simple audio generation, deliberate-thinking, TODO, group mute and active memory. Output is not limited to text — replies can render as text, images or voice, and vision-capable models accept multimodal image input. Per-group presets let each room carry its own prompt, render template and world book, while resource management enforces cooldowns, concurrency limits and call caps so a busy group shares one API budget predictably. It suits Koishi bot developers, self-hosters and community operators. It is a plugin, not an app: there is no standalone client, and cost and rate limits come from the model platforms whose keys you supply. Documentation was moved to a VitePress site with structured guides for setup, model platforms, agent tools and presets, and the agent sections were refreshed to match the shipped tool set.
Behind the Verdict
ChatLuna's real value is consolidation. A Koishi operator who wants DeepSeek for one group, Claude for another and Gemini for a third would otherwise wire up three separate provider plugins; ChatLuna replaces that with one install and a model-platform selector covering OpenAI, Google Gemini, Claude, DeepSeek, Qwen, Dify, ChatGLM, iFlytek Spark, ERNIE, Hunyuan, Moonshot, Ollama, RWKV and OpenAI-compatible endpoints. Because it is built on LangChain and exposes a model-service API, the layer is reusable — long-term memory, knowledge-base and custom renderers are plugins on top of the same interface rather than forks of the chat code. Where it goes beyond chat wrappers is Agent mode. Installing the agent plugins hands the model a sandbox and a broad tool set: MCP support, web search, page browsing, arbitrary URL requests, OpenAPI requests, file read/write, scheduled tasks, Koishi shell command execution, text-to-image, simple audio generation, deliberate-thinking and TODO tools, plus a group-mute moderation tool and an active memory tool. That is a genuinely wide surface for a community bot, and it is assembled for you rather than hand-rolled. Output handling is the second differentiator. Replies render as text, images or voice, image-capable models accept multimodal input, and presets give each group or room its own prompt, render template and world book. Resource management adds global cooldowns, concurrency limits and call caps — the practical answer to a shared API budget in a busy group — and content filtering hooks into Koishi's censor service. The honest constraints: ChatLuna needs a Koishi host, so it is not a product you can sign up for and use tomorrow without that stack. Model access depends on keys you supply, and cost, quota and availability come from those upstream platforms, not from ChatLuna. Agent quality is bounded by the model's tool-calling reliability and context window — a model that calls tools poorly will do so here too. The documentation is Chinese-first and community-maintained, and the guides were only recently restructured onto a VitePress site. Setup means filling in Koishi plugin configuration rather than running a standalone binary. If you are on Koishi, those are fair trades; if you are not, a standalone client is the shorter path.
Researching Chatluna? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Chatluna actually fits — and what changes day-one when you adopt it.
You install ChatLuna in your Koishi instance, add the DeepSeek and Gemini model-platform plugins, drop in the API keys you already have, and point two different groups at two different models from the same configuration.
Outcome: One plugin install replaces several single-provider plugins, and each group gets the model that suits it without a second codebase.
You install the agent plugins, enable MCP, web search and page browsing, set a per-group preset with its own prompt and world book, then cap concurrency and call counts for the room.
Outcome: The bot answers live questions with current information and runs multi-step tool tasks, while a shared API budget stays bounded by the cooldown and call limits you set.
You build against ChatLuna's exposed model-service API instead of calling providers directly, and register a custom tool or renderer as a Koishi plugin.
Outcome: Your plugin works across every model platform ChatLuna supports without you writing per-provider adapters, and it inherits preset, memory and rendering behavior.
Use Cases
- Route different Koishi groups to different models (Gemini, Claude, DeepSeek) from one plugin install
- Give a bot an agent that reads files, calls APIs and runs scheduled tasks through sandboxed tools
- Attach external MCP tool servers to your Koishi bot's model loop
- Let a bot answer live questions with the built-in web search and page browsing tools
- Render model replies as voice messages for community channels
- Configure per-group presets with distinct prompts, world books and render templates
- Add a knowledge-base or long-term-memory layer on top of the exposed model-service API
- Build custom model adapters, tools or renderers as Koishi plugins
Models Under the Hood
as of 2026-10-18
Limitations
- ChatLuna is a plugin, not an application: it requires a Koishi installation and configuration before it does anything.
- Model access depends on the API keys you supply and on the availability and terms of the upstream platforms, so cost and rate limits come from those providers rather than from ChatLuna.
- Agent capability is bounded by the underlying model's tool-use reliability and context window — a model that calls tools poorly will do so inside ChatLuna too.
- The docs are Chinese-first and community-maintained, and were only recently restructured onto a VitePress site.
- Setup means filling in Koishi plugin configuration rather than installing a standalone binary.
as of 2026-10-16
Verification history
We have re-verified Chatluna 9 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
Showing the 6 most recent of 9 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where Chatluna's pricing actually pencils out — and where peers do it cheaper.
ChatLuna itself is released under CC-BY-SA-4.0 at no cost, so the economic comparison is against other Koishi chat plugins (also free, but usually single-provider) and against standalone clients like Open WebUI or Jan — those are free too, but you trade away ChatLuna's Koishi-native group presets and agent tool sandbox. Your actual spend is set by whichever model platform's API keys you plug in.
Setup time & first value
How long it actually takes to get something useful out of Chatluna — broken out by persona, not the marketing-page minute.
Koishi bot developers who already have a running instance: minutes to install ChatLuna and a model-platform plugin, plus the time to paste API keys and pick a model. Community operators adding per-group presets: under an hour. Plugin authors building on the model-service API: hours, depending on how much custom tooling you write.
Switching to or from Chatluna
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From single-provider Koishi chat plugins: install ChatLuna plus the matching model-platform plugin, port your prompt into a preset, and retire the old plugin.
- →From a hand-rolled LangChain script outside Koishi: move provider credentials into ChatLuna's model-platform config and rebuild the tool logic as agent plugins or MCP servers.
- →From ChatHub: ChatLuna is the renamed successor (the repo history shows 'rename chathub to chatluna'), so migrate existing chathub configuration to ChatLuna naming and presets.
- ↗To Open WebUI: export your prompts as system messages in Open WebUI, but you lose Koishi group presets, the renderer pipeline and the agent tool sandbox.
- ↗To Jan: copy prompt and world-book content into Jan's assistant config, accepting that Jan is a desktop client with no Koishi group or channel integration.
- ↗To a custom LangChain service: reuse ChatLuna's exposed model-service concepts as a reference, but rebuild presets, resource limits and rendering yourself.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “Chatluna”, and we withheld 6: 6 could not be judged, because “Chatluna” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Chatluna.
Official links
Tools that pair well with Chatluna
Common stack mates teams adopt alongside Chatluna, with the specific reason each pairing earns its keep.
CopilotKit
Open-source React framework for agent chat, generative UI, and shared state on any AG-UI backend.
Chatbox
Chatbox is a multi-model AI chat client with a local knowledge base and a Work Mode desktop agent that runs code and finishes tasks.
Rikkahub
Free, open-source Android AI chat client with a proot Linux agent workspace and MCP tool calling, using your own API keys.
Featured Head-to-Head Comparisons
Chatluna vs Presto Voice
Choose Presto Voice if you run a QSR chain and need a turnkey drive-thru AI that boosts revenue via upselling (up to 6% lift) and integrates with existing POS/headsets. Choose Chatluna if you're a Koishi developer wanting a free, extensible plugin to connect multiple LLMs and build agentic chatbots. These tools serve completely different domains, so selection hinges on your business vertical—restaurant operations vs. chatbot development.
Chatluna vs Locus Robotics
There is no decision here. Locus Robotics is a capital-and-subscription warehouse automation purchase — physical AMRs, WMS integrations, and a RaaS contract — bought by operations leaders at 3PLs and fulfillment centers. Chatluna is a free Koishi plugin that gives a bot multi-LLM chat, tool calling, and MCP. If you run a warehouse and need picking throughput, Locus is the shortlist candidate and Chatluna is irrelevant. If you run a Koishi bot and want model-switching chat, Chatluna is the answer and Locus is not a product you can install. Pick the one that matches your problem; there is no overlap to weigh.
Chatluna vs Truleo
Choose Truleo if you're a law enforcement agency needing to unify siloed data for lead generation, jail call analysis, and report automation. Choose ChatLuna if you're a Koishi developer wanting a free, multi-model plugin with agentic capabilities. They serve entirely different markets and use cases.
Alternatives to Chatluna
View all
CopilotKit
Open-source React framework for agent chat, generative UI, and shared state on any AG-UI backend.
Frequently Asked Questions
Used Chatluna? Help shape our editorial sentiment research.