Which model should I use?
- Claude Opus 5 (
claude-opus-5) — maximum intelligence for difficult, long-horizon work. - GPT-5.6 (
gpt-5.6) — fast on complex tasks. - Gemini 3.5 Flash (
gemini-3.5-flash) — fast for simpler tasks. - MiniMax M3 (
minimax-m3, default) — cheapest for simple and high-volume tasks.
How do I get the live browser URL?
The V4 run’sbrowser.ready event contains live_view_url. Embed it in an iframe or open it in a browser.
browser.ready appears if you need the URL while the run is still active. See Human in the loop for a complete flow.
Getting blocked by a website
Stealth and proxies are active by default. If you’re still getting blocked:- Use a profile with logged-in cookies to bypass login walls.
- Try a different proxy country to match the target region.
Rate limited (429 errors)
The SDK auto-retries 429 responses with exponential backoff. If persistent, you may need more concurrent sessions — contact support.V2 or V4 — which should I use?
Use V4 for difficult tasks where accuracy matters. It supports:- Run-focused API with a cheap status polling endpoint
- Conversation sessions with follow-ups
- Persistent workspaces and turn-scoped file attachments
- Incremental events for custom UIs and monitoring
- Per-run cost totals, cost caps, and optional judgement