Open WebUI 리트윗함
Open WebUI
346 posts
Open WebUI 리트윗함
the headline in openclaw 3.24 for me: Open WebUI finally works like a real frontend.
before this, the OpenAI-compatible layer was the weak link. tool_call_id duplicates, strict field errors on non-native providers, prompt_cache breakage - you'd debug the bridge instead of the agent.
3.22 started fixing it. 3.24 finishes it. the gateway now runs the same agent codepath, so routing, tools, and sub-agents all fire normally behind the HTTP surface.
setup in docker:
# openclaw gateway config
gateway.http.endpoints.chatCompletions.enabled=true
# docker-compose.yml (Open WebUI)
services:
open-webui:
environment:
- OPENAI_API_BASE_URL=http://host.docker.internal:18789/v1
- OPENAI_API_KEY=your-openclaw-key
if Open WebUI has a persistent volume from a previous run, env changes might not apply - update via Admin Panel instead.
Open WebUI picks up your top-level agent targets: openclaw/default, openclaw/scout, etc. select one, chat. tools and sub-agents fire on the backend. you just get a UI for it now.
the best agent runtime isn't the one with the most features. it's the one that lets you choose how you talk to it.
Peter Steinberger 🦞@steipete
New @openclaw beta is out with better MS Teams integration, @OpenWebUI and more!
English
@dennysentinel @steipete @openclaw We don't depend on LiteLLM, so Open WebUI itself isn't affected. With that being said, some deployments that use LiteLLM separately in their stack may want to double check on their end. 💪
English
English
Open WebUI 리트윗함
Open WebUI 리트윗함
OpenClaw 2026.3.24 🦞
🔌 Improved OpenAI API: talk to sub-agents with @openwebui
🎛️ Skill & tool management Control UI
🎨 Slack interactive reply buttons
💅 Native Microsoft Teams
🧵 Smart Discord auto-thread naming
Any client. Any model. One runtime. github.com/openclaw/openc…
English
Open WebUI 리트윗함
🤝 Hello native subagents from @openclaw inside @OpenWebUI. We revamped our OpenAI compatible API to give sub-agent level chat support in todays build. Image and speech is coming soon 💅
English
Open WebUI 리트윗함
Private AI infra is becoming the default.
GPU inference → API layer → UI
Here’s a practical guide using Open WebUI + vLLM on Hot Aisle (MI300X) to build a private ChatGPT-style system:
hotaisle.xyz/blog/chatxyz-o…
@HotAisle @AIatAMD @AMD @OpenWebUI @vllm_project
English
Open WebUI 리트윗함
We have to acknowledge that @OpenWebUI + open terminal with @MistralAI 4 small is a game changer for AI sovereignty. Well done to both of you.
English
Open WebUI 리트윗함
I've drop-in solution called Ovllm. It's essentially an Ollama-style wrapper, but for vLLM instead of llama.cpp. It's still a work in progress; it downloads models directly from Hugging Face. @vllm_project @OpenWebUI github.com/FearL0rd/Ovllm OpenWebUi Compatible
English
Open WebUI 리트윗함
Open WebUI 리트윗함
The Qwen 3.5 small model series is now available
ollama run qwen3.5:9b
ollama run qwen3.5:4b
ollama run qwen3.5:2b
ollama run qwen3.5:0.8b
All models support native tool calling, thinking, and multimodal capabilities in Ollama.
Qwen@Alibaba_Qwen
🚀 Introducing the Qwen 3.5 Small Model Series Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B ✨ More intelligence, less compute. These small models are built on the same Qwen3.5 foundation — native multimodal, improved architecture, scaled RL: • 0.8B / 2B → tiny, fast, great for edge device • 4B → a surprisingly strong multimodal base for lightweight agents • 9B → compact, but already closing the gap with much larger models And yes — we’re also releasing the Base models as well. We hope this better supports research, experimentation, and real-world industrial innovation. Hugging Face: huggingface.co/collections/Qw… ModelScope: modelscope.cn/collections/Qw…
Suomi
Open WebUI 리트윗함
Me realising I can run these models locally on my M1 MacBook Air for free
Qwen@Alibaba_Qwen
🚀 Introducing the Qwen 3.5 Small Model Series Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B ✨ More intelligence, less compute. These small models are built on the same Qwen3.5 foundation — native multimodal, improved architecture, scaled RL: • 0.8B / 2B → tiny, fast, great for edge device • 4B → a surprisingly strong multimodal base for lightweight agents • 9B → compact, but already closing the gap with much larger models And yes — we’re also releasing the Base models as well. We hope this better supports research, experimentation, and real-world industrial innovation. Hugging Face: huggingface.co/collections/Qw… ModelScope: modelscope.cn/collections/Qw…
English
How Okta Deploys and Secures AI for its Employees
okta.com/blog/ai/okta-a…
English
Open WebUI 리트윗함
Running local models shouldn’t be complicated.
OpenWebUI now auto-detects Docker Model Runner for a self-hosted setup in minutes.
Enable Model Runner, start OpenWebUI, and your models are live in the browser. No extra config, no cloud required.
📖 bit.ly/46ucxE1
English
Open WebUI 리트윗함
Open WebUI 리트윗함
I wonder if @OpenWebUI has any plan to support Responses API directly 🤔
English
Responses API from OpenAI is so much better! Everyone should migrate to it ASAP!
Note: it works really well on Azure too 🚀
English