diff --git a/clawspring/README.md b/clawspring/README.md
index b248cd12..339ab855 100644
--- a/clawspring/README.md
+++ b/clawspring/README.md
@@ -375,6 +375,7 @@ Claude Code is a powerful, production-grade AI coding assistant — but its sour
 | **Zhipu (GLM)** | `glm-4-flash` | 128k | Free tier available | `ZHIPU_API_KEY` |
 | **DeepSeek** | `deepseek-chat` | 64k | Strong coding | `DEEPSEEK_API_KEY` |
 | **DeepSeek** | `deepseek-reasoner` | 64k | Chain-of-thought reasoning | `DEEPSEEK_API_KEY` |
+| **MiniMax** | `MiniMax-M3` | 1M | Long-context coding and agent tasks | `MINIMAX_API_KEY` |
 
 ### Open-Source (Local via Ollama)
 
@@ -542,6 +543,16 @@ clawspring --model deepseek/deepseek-chat
 clawspring --model deepseek/deepseek-reasoner
 ```
 
+### MiniMax
+
+Get your API key at [platform.minimax.io](https://platform.minimax.io/docs).
+
+```bash
+export MINIMAX_API_KEY=...
+
+clawspring --model MiniMax-M3
+```
+
 ---
 
 ## Usage: Open-Source Models (Local)
@@ -729,6 +740,7 @@ clawspring --model qwen:qwen-max
 | `qwen`, `qwq-` | qwen |
 | `glm-` | zhipu |
 | `deepseek-` | deepseek |
+| `minimax-` | minimax |
 | `llama`, `mistral`, `phi`, `gemma`, `mixtral`, `codellama` | ollama |
 
 ---
@@ -869,6 +881,7 @@ export MOONSHOT_API_KEY=sk-...       # Kimi
 export DASHSCOPE_API_KEY=sk-...      # Qwen
 export ZHIPU_API_KEY=...             # Zhipu GLM
 export DEEPSEEK_API_KEY=sk-...       # DeepSeek
+export MINIMAX_API_KEY=...           # MiniMax
 ```
 
 ### Method 2: Set Inside the REPL (persisted)
@@ -881,6 +894,7 @@ export DEEPSEEK_API_KEY=sk-...       # DeepSeek
 /config qwen_api_key=sk-...
 /config zhipu_api_key=...
 /config deepseek_api_key=sk-...
+/config minimax_api_key=...
 ```
 
 Keys are saved to `~/.clawspring/config.json` and loaded automatically on next launch.
diff --git a/clawspring/providers.py b/clawspring/providers.py
index 9575b1c8..0144217c 100644
--- a/clawspring/providers.py
+++ b/clawspring/providers.py
@@ -9,6 +9,7 @@
   qwen       — Alibaba DashScope (qwen-max, qwen-plus, ...)
   zhipu      — Zhipu GLM (glm-4, glm-4-plus, ...)
   deepseek   — DeepSeek (deepseek-chat, deepseek-reasoner, ...)
+  minimax    — MiniMax (MiniMax-M3)
   ollama     — Local Ollama (llama3.3, qwen2.5-coder, ...)
   lmstudio   — Local LM Studio (any loaded model)
   custom     — Any OpenAI-compatible endpoint
@@ -99,6 +100,15 @@
             "deepseek-chat", "deepseek-coder", "deepseek-reasoner",
         ],
     },
+    "minimax": {
+        "type":       "openai",
+        "api_key_env": "MINIMAX_API_KEY",
+        "base_url":   "https://api.minimax.io/v1",
+        "context_limit": 1000000,
+        "models": [
+            "MiniMax-M3",
+        ],
+    },
     "ollama": {
         "type":       "ollama",
         "api_key_env": None,
@@ -145,6 +155,7 @@
     "qwen-plus":                (0.4,   1.2),
     "deepseek-chat":            (0.27,  1.1),
     "deepseek-reasoner":        (0.55,  2.19),
+    "MiniMax-M3":                (0.6,   2.4),
     "glm-4-plus":               (0.7,   0.7),
 }
 
@@ -161,6 +172,7 @@
     ("qwq-",          "qwen"),
     ("glm-",          "zhipu"),
     ("deepseek-",     "deepseek"),
+    ("minimax-",      "minimax"),
     ("llama",         "ollama"),
     ("mistral",       "ollama"),
     ("phi",           "ollama"),