Runner now offers Claude Opus 4.8 — Anthropic's most capable model — so you can reach for extra reasoning power whenever a task calls for it. This release also makes action cards simpler to scan and approve, and keeps your conversations clean and easy to follow.
What changed:
- Models: Choose Claude Opus 4.8 for your most demanding tasks
- Action cards: Cleaner, simpler cards that are easier to scan and approve at a glance
- Action cards: Sending to a WhatsApp group now shows the group name on the card, so you can confirm exactly where your message is going
- Settings: Simpler, clearer browser profile controls
- Conversations: A tidier chat — internal routing labels no longer appear while Runner is replying
- Fixed an issue where a generated file, like a PDF, could fail to open from a conversation
- Fixed an issue where a saved automation could fail to run because of an outdated model selection
- Fixed the placement of the back button on the connect screen during setup
- General stability improvements
Why it matters: The right model for the moment makes Runner more capable on hard problems, while clearer action cards and a tidier chat make it easier to see what Runner is doing and approve it with confidence.