Roadmap — Next Features

Prioritized list of planned features for the gpt4free browser extension. ✅ = shipped in v0.1 · 🚧 = in progress · ⬜ = planned

v0.2 — Accounts & Sync (current)

⬜ Conversation search & history browser — full-text search across past chats
⬜ Prompt library — save / reuse / share prompt templates with variables
⬜ Provider picker per chat — override the provider per conversation, not just globally
⬜ Regenerate & edit — re-run the last exchange, edit a sent message
⬜ Multi-model compare — send one prompt to two models side by side
⬜ Token/usage stats — per-conversation token estimates, usage via members API

⬜ Voice input — Web Speech API dictation into the panel
⬜ Vision / image attach — paste or drop images for vision models (m.vision)
⬜ Text-to-speech — read answers aloud via /v1/audio endpoints
⬜ YouTube / PDF context — transcript & document ingestion as page context
⬜ LaTeX rendering — KaTeX in markdown output

⬜ Page Q&A mode — auto-RAG over the current page, no checkbox needed
⬜ Selection inline actions — floating "Ask g4f" bubble on text selection
⬜ Custom context menus — user-defined right-click actions from the prompt library
⬜ Keyboard-first mode — command palette (Ctrl+K) for actions & model switching
⬜ Export — copy conversation as Markdown / JSON, download images

MCP integration: expose extension page-context as an MCP tool for the g4f server
Shared team prompt packs via secret workspace
Side-panel notebook mode (multi-turn scratchpad with cells)
Auto-model routing (cheap model for short prompts, big model for code)