Switch inference to Tinker production endpoint (docs snippet)

Fix docs modal: move markup before script so listeners attach

Header 'How to use' button opens docs popover with API snippet

Promote playground link to centered hero CTA

Add 'Try the Playground' link to header

ChatGPT-style image upload; clean UI (no extra hints)

Fix vision payload: URL from imgUrl input (not file upload)

Final UI cleanup: remove NEW badge, clean image button, update hint text

Clean UI: hide provider/temp/max-tokens, round arrow send, remove backend hint

Make image input optional file upload; clean UI

Fix empty message validation; clean vision message flow

Select 'together' as default in dropdown; add selected

Apply Thinking Machines Lab aesthetic + logo; vision input

Add vision message support; image URL input in composer

Switch frontend to /gradio_api/call/v2 protocol; join before SSE

Silence harmless ValueError: Invalid file descriptor on Space shutdown

Split payload into positional args for queue/join

Replace gradio JS client with raw SSE protocol; tighten stream_every

Add Inkling chat app: gradio.Server backend + HF Inference Providers stream