Every provider claude-multi supports.
claude-multi routes Claude Code to the provider that fits your budget and workload. Each provider exposes a native Anthropic-compatible endpoint, so setup is a single command. Pick one to see setup steps, pricing, and where it fits best.
GLM-5.3 Coding Plan Provider for Claude Code
glmRun Claude Code with GLM-5.3, GLM-5.3-Flash, and GLM-5-Turbo via z.ai Coding Plan subscription. Full Anthropic API compatibility, up to 1M context, thinking mode enabled.
claude-multi add glmCoding Plan subscription via z.aiMiniMax M3 Provider for Claude Code
minimaxRun Claude Code with MiniMax-M3 via minimax.io. 1M token context window, 512K max output, native multimodal support, one model across every role.
claude-multi add minimaxPay-per-token via minimax.ioDeepSeek Provider for Claude Code
deepseekRun Claude Code with DeepSeek-V4-Pro and DeepSeek-V4-Flash via deepseek.com. 1M context, thinking mode, pay-per-token pricing, no subscription.
claude-multi add deepseekPay-per-token via deepseek.comXiaomi MiMo Provider for Claude Code
mimoRun Claude Code with MiMo-V2.5-Pro and MiMo-V2.5 via xiaomimimo.com. 1T parameter MoE model, 1M context, pay-per-token pricing with a token plan option.
claude-multi add mimoPay-per-token via xiaomimimo.comMoonshot Kimi provider for Claude Code
kimiRun Claude Code with Kimi K2.7 Code, K2.6, and K2.5 via moonshot.ai. Step-by-step setup, model specs, pricing, and when to pick Kimi over DeepSeek or GLM.
claude-multi add kimiPay-per-token via moonshot.aiAlibaba Qwen Provider for Claude Code
qwenRun Claude Code with Qwen3-Coder-Next, Qwen3-Coder-Plus, and Qwen3-Coder-Flash via Alibaba DashScope. Three model tiers, pay-per-token pricing, 128K context.
claude-multi add qwenPay-per-token via Alibaba DashScopeTemplate reference
Each template pre-fills ANTHROPIC_BASE_URL, model mappings, and related env vars into your instance's settings.json. You only need to supply your API key. All providers expose an Anthropic-compatible API, so claude-multi points the same Claude Code binary at whichever endpoint you pick.
| Template | Display name | Endpoint | Opus model | Sonnet/Haiku model |
|---|---|---|---|---|
| glm | GLM Coding Plan | api.z.ai | glm-5.3[1m], glm-5.3-flash | glm-5-turbo |
| minimax | MiniMax | api.minimax.io | MiniMax-M3 | MiniMax-M3 |
| deepseek | DeepSeek | api.deepseek.com | deepseek-v4-pro[1m] | deepseek-v4-flash |
| mimo | Xiaomi MiMo | api.xiaomimimo.com | mimo-v2.5-pro | mimo-v2.5 |
| mimo-token | Xiaomi MiMo (Token Plan) | token-plan-cn.xiaomimimo.com | mimo-v2.5-pro | mimo-v2.5 |
| kimi | Moonshot Kimi | api.moonshot.ai | kimi-k2.7-code | kimi-k2.6, kimi-k2.5 |
| qwen | Alibaba Qwen | dashscope-intl.aliyuncs.com | qwen3-coder-next | qwen3-coder-flash |
| qwen-coding | Alibaba Qwen Coding Plan | coding-intl.dashscope.aliyuncs.com | qwen3-coder-next | qwen3-coder-flash |
Pay per token vs. subscription
Some providers offer two access models, a pay-per-token API and a subscription/coding plan, each with a different base URL. Use the right template for your account type:
| Provider | Pay-per-token template | Subscription template |
|---|---|---|
| Xiaomi MiMo | mimo | mimo-token (regional URL) |
| Alibaba Qwen | qwen | qwen-coding |
| GLM (Z.ai) | no Anthropic URL | glm (coding-plan-only) |
| MiniMax | minimax | same URL, different key type |
| Moonshot Kimi | kimi | pay per token only |
| DeepSeek | deepseek | pay per token only |
- GLM: the Anthropic-compatible endpoint (
api.z.ai/api/anthropic) is exclusive to the Z.ai Coding Plan subscription. The metered API only exposes an OpenAI-compatible URL. - MiMo Token Plan: defaults to the CN regional endpoint. If your subscription is SG or EU, update
ANTHROPIC_BASE_URLwith the endpoint shown in your subscription console. - Kimi: no subscription plan, strictly pay per token at
api.moonshot.ai. - MiniMax: both plans use the same
api.minimax.ioendpoint; the API key type determines which quota is consumed. MiniMax-M3 has a 1M token context window and accepts text, image, and video inputs.
Not sure which provider to pick?
The getting started guide walks through installing claude-multi and adding your first instance. It takes about two minutes to get going.