[3.8.1]
Update open-webui to 0.13.3
Full Changelog
Security Advisory: This release contains security fixes. We strongly recommend updating all deployments as soon as possible to have the latest security fixes in place. These fixes are not listed in this changelog; we will publish the security advisories for them immediately following CVE assignment. Advisories
Database Migrations: This release includes database schema changes; we strongly recommend backing up your database and all associated data before upgrading in production environments. If you are running a multi-worker, multi-server, or load-balanced deployment, all instances must be updated simultaneously, rolling updates are not supported and will cause application failures due to schema incompatibility.
Preload files off by default. With native function calling, the default, attached files are no longer added to the conversation automatically, and models read them with their file tools instead. Models without tool support or without the Files built-in tool can no longer see attached files until you turn on Preload files in the model's Capabilities. Legacy function calling, and API requests that do not come from a chat in the web UI, keep the old behavior. #33815
Chat over plain HTTP. Opening Open WebUI over plain HTTP at a LAN or Docker address now loads new and existing chats instead of staying on a spinner with "crypto.randomUUID is not a function". #33919
Agents in automations. The automation editor now lets you choose a terminal and working folder, then select a coding agent available there. Scheduled runs use that saved terminal and folder, and check that the terminal and agent are still available before starting. Agent automations send their results to a chat; channel destinations remain available for ordinary models. Commit
Coding agents as models. On an Open Terminal that offers agents, coding agents such as Claude Code, Codex and Pi now appear in the model picker. You can chat with them or call them by voice, choose their reasoning effort and service tier, and approve their tool requests in the chat. Each conversation uses one agent, and agent chats cannot be regenerated, continued or edited and resent. #33843, #33860, #33880, #33857
Much cheaper tool chats. Chats that use tools now cost much less. Citation instructions, memory recall, the skills list and attached knowledge are no longer rewritten into the system prompt or your message on every tool round, so the start of the conversation stays the same between requests and providers can serve it from their prompt cache. #33852
Side chats are saved. Side chats, which were always temporary in 0.13.0, are now saved as regular chats and appear in the chat list. They stay temporary only when you turn on Temporary Chat in the side chat or your admin enforces temporary chats.
Ollama broken tool calls. Chats with Ollama models no longer get stuck after a tool call with cut-off, rejected or empty arguments. #33697, #33671
DeepSeek thinking mode with skills, knowledge and memories. These chats no longer fail with a missing reasoning error when the connection's Provider is set to "llama.cpp". DeepSeek thinking mode still requires that setting; connections using "Default" are unchanged. #33950
