成功在本地部署openclaw使用本地模型并可以exec执行任务tool执行成功配套软件版本1.ollama 0.18.2 # 安装在windows下本机显卡 NVIDIA GeForce RTX 4050 Laptop GPU 6G vram2.openclaw v2026.03.18 #安装在 WSL2 cpu4核 内存4G Ubuntu系统3.模型扩窗 链接4.contextWindow 值越大启动同一个模型占用的内存越多视显存情况调整本文是128000挺大如果显存小可调成65536或更小windows用户文件夹下.wslconfig的内容如下[wsl2]memory4GB# 内存swap2G#虚拟内存processors4# 核心数[experimental]#autoMemoryReclaimgradual # 开启自动回收内存可在 gradual, dropcache, disabled 之间选择networkingModemirrored# 开启镜像网络dnsTunnelingtrue# 开启 DNS Tunnelingfirewalltrue# 开启 Windows 防火墙autoProxytrue# 开启自动同步代理#sparseVhdtrue # 开启自动释放 WSL2 虚拟硬盘空间测试ollamacurlhttp://localhost:11434/v1/chat/completions-HContent-Type: application/json-d{model:qwen2.5:latest-32k,messages:[{role:user,content:Hello!}]}openclaw.json{meta:{lastTouchedVersion:2026.3.13,lastTouchedAt:2026-03-18T02:55:54.422Z},wizard:{lastRunAt:2026-03-16T08:58:14.320Z,lastRunVersion:2026.3.13,lastRunCommand:doctor,lastRunMode:local},auth:{profiles:{qwen-portal:default:{provider:qwen-portal,mode:oauth}}},models:{providers:{qwen-portal:{baseUrl:https://portal.qwen.ai/v1,apiKey:qwen-oauth,api:openai-completions,models:[{id:coder-model,name:Qwen Coder,reasoning:false,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:128000,maxTokens:8192},{id:vision-model,name:Qwen Vision,reasoning:false,input:[text,image],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:128000,maxTokens:8192}]},ollama:{baseUrl:http://127.0.0.1:11434/v1,apiKey:ollama,api:openai-completions,models:[{id:deepseek-r1:latest,name:DeepSeek-R1-0528-Qwen3-8B (Ollama),reasoning:true,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:32768,maxTokens:4096},{id:qwen2.5:latest,name:Alibaba Qwen2.5 (Ollama),reasoning:false,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:128000,maxTokens:8192},{id:qwen2.5:latest-32k,name:Alibaba Qwen2.5:latest-32k (Ollama),reasoning:false,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:128000,maxTokens:8192}]}}},agents:{defaults:{model:{primary:ollama/qwen2.5:latest-32k,fallbacks:[]},models:{qwen-portal/coder-model:{alias:qwen},qwen-portal/vision-model:{},ollama/deepseek-r1:latest:{alias:deepseek},ollama/qwen2.5:latest:{alias:qwen-local},ollama/qwen2.5:latest-32k:{alias:qwen-local-32k},ollama/qwen2.5:1.5b:{alias:qwen2.5:1.5b}},workspace:/home/tj/.openclaw/workspace,compaction:{mode:safeguard},maxConcurrent:4,subagents:{maxConcurrent:8}},list:[{id:main,model:ollama/qwen2.5:latest-32k}]},tools:{profile:full},messages:{ackReactionScope:group-mentions},commands:{native:auto,nativeSkills:auto,restart:true,ownerDisplay:raw},session:{dmScope:per-channel-peer},hooks:{internal:{enabled:true,entries:{bootstrap-extra-files:{enabled:true}}}},gateway:{port:18789,mode:local,bind:loopback,controlUi:{allowedOrigins:[http://localhost:18789,http://127.0.0.1:18789],allowInsecureAuth:true,dangerouslyDisableDeviceAuth:true},auth:{mode:token,token:xxxxxxxxxxx},tailscale:{mode:off,resetOnExit:false},nodes:{denyCommands:[camera.snap,camera.clip,screen.record,contacts.add,calendar.add,reminders.add,sms.send]}},skills:{install:{nodeManager:npm}},plugins:{entries:{qwen-portal-auth:{enabled:true},imessage:{enabled:true}}}}关于不干活问题亲测qwen2.5:latest-32k不能稳定干活经常找不着本机目录的准确位置也导致无法正常执行本机操作。根本原因是模型不够用换成gag0/qwen35-opus-distil:27bcarstenuhlig/omnicoder-9b:latest等问题解决。安装命令如下1.显存超过32G可用gag0/qwen35-opus-distil:27bqwen3.5:27bollama launch openclaw--modelgag0/qwen35-opus-distil:27b2.显存没有32G可用carstenuhlig/omnicoder-9b:latestqwen3.5:9bollama launch openclaw--modelcarstenuhlig/omnicoder-9b:latest一个能干活的本地模型配置案例{wizard:{lastRunAt:2026-03-30T03:15:09.551Z,lastRunVersion:2026.3.28,lastRunCommand:onboard,lastRunMode:local},auth:{profiles:{ollama:default:{provider:ollama,mode:api_key}}},models:{mode:merge,providers:{ollama:{baseUrl:http://127.0.0.1:11434,apiKey:OLLAMA_API_KEY,api:ollama,models:[{id:glm-4.7-flash,name:glm-4.7-flash,reasoning:false,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:128000,maxTokens:8192},{id:carstenuhlig/omnicoder-9b:latest,name:carstenuhlig/omnicoder-9b:latest,reasoning:false,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:262144,maxTokens:8192},{id:gag0/qwen35-opus-distil:27b,name:gag0/qwen35-opus-distil:27b,reasoning:false,input:[text],cost:{input:0,output:0,cacheRead:0,cacheWrite:0},contextWindow:262144,maxTokens:8192}]}}},agents:{defaults:{model:{primary:ollama/carstenuhlig/omnicoder-9b:latest},models:{ollama/carstenuhlig/omnicoder-9b:latest:{}},workspace:/home/tj/.openclaw/workspace}},tools:{profile:coding,alsoAllow:[ollama_web_search,ollama_web_fetch]},commands:{native:auto,nativeSkills:auto,restart:true,ownerDisplay:raw},session:{dmScope:per-channel-peer},hooks:{internal:{enabled:true,entries:{boot-md:{enabled:true},bootstrap-extra-files:{enabled:true},command-logger:{enabled:true},session-memory:{enabled:true}}}},gateway:{port:18789,mode:local,bind:loopback,controlUi:{allowInsecureAuth:true},auth:{mode:token,token:XXXXXXXXXX},tailscale:{mode:off,resetOnExit:false},nodes:{denyCommands:[camera.snap,camera.clip,screen.record,contacts.add,calendar.add,reminders.add,sms.send]}},skills:{install:{nodeManager:npm}},meta:{lastTouchedVersion:2026.3.28,lastTouchedAt:2026-03-30T03:15:09.563Z}}体验一下启动openclaw gateway--port18789在http://127.0.0.1:18789/chat聊天中输入安装desearch-web-search提示配置api key 时按提示到https://console.desearch.ai注册账号获取key在聊天中输入API key 是 dsr_xxxxxxxxxxxxxxxxxxxxx等openclaw自动配置好后就可以网络搜索了。免费搜索用安装web-search-free补充安装 ollama/openclaw-web-search 联网插件ollama/openclaw-web-search 是 OpenClaw 与 Ollama 实现联网搜索的核心插件无需手动下载安装包通过 OpenClaw 官方命令即可一键完成安装步骤如下打开终端确保当前终端拥有管理员权限Windows 右键 PowerShell 选择“以管理员身份运行”macOS 命令前加sudo避免插件安装过程中出现权限不足报错。执行插件安装命令直接调用 OpenClaw 插件管理功能完成安装命令如下openclaw pluginsinstallollama/openclaw-web-search安装完成后执行验证命令确认插件安装成功且处于激活状态openclaw plugins list若输出结果中包含“ollama/openclaw-web-search”且状态为“loaded”则插件安装成功若状态为“disabled”执行以下命令激活插件openclaw pluginsenableollama/openclaw-web-search