fix: recognize image input for GPT-6.2 models routed through ClawRouter When neither the selected endpoint nor the model row declares input, ClawRouter's fallback classified gpt-6.2-sol and gpt-6.2-astra as text-only. Add both upstream IDs and cover them in the fallback test.
fix: recognize image input for GPT-6.2 models routed through ClawRouter
Find Your Interest
- AI Agents Could Drive the Next Open Source Software Revolution
- The Web is for People
- 2024 Year in Review & Looking Ahead
- The open-source movement wants Jensen Huang back
- This Week’s Awesome Tech Stories From Around the Web (Through November 14)
- How Fin Doubled Its Engineering Capabilities With AI Agents
- Disney picks Adobe as its new AI and price-hike partner
- Developers, Your Next Customer Is an AI Agent
- Jesse Eisenberg "experiments" with AI to make sure it's still bad at writing
- This Week’s Awesome Tech Stories From Around the Web (Through November 7)





