Launching today
Valefront
One API that picks the right AI model for every prompt
2 followers
One API that picks the right AI model for every prompt
2 followers
Valefront is an LLM router. Point the OpenAI or Anthropic SDK at llm.valefront.com, send "model": "auto", and each request goes to a model that fits its task, quality bar, cost and latency, with automatic fallbacks. Pay as you go, no markup.
Free Options
Launch Team
Render Sandboxes: a computer for every agentGive every agent its own machine. Starts in under a second
Promoted
Hi Product Hunt, I'm Steve, I build Valefront with a small team in Da Nang, Vietnam.
We kept hardcoding one model per feature, then paying premium prices for prompts a small model handles fine, and scrambling when a provider slowed down. Valefront is the layer we wanted: one endpoint, and you choose a routing strategy (cheapest, fastest, best, balanced, or your own rules) instead of a model name.
What you get today:
OpenAI- and Anthropic-compatible API at llm.valefront.com, so Claude Code, Codex, OpenCode and Pi work by changing the base URL
routing with automatic fallbacks
per-key budgets, PII redaction, and a usage dashboard with cost per request
prepaid credit at the providers' official per-token prices, no subscription
1M free tokens to try it, 4M more after your first top-up
bring your own keys for other providers at no charge
It launched yesterday, so it's still rough in places. There's a "Report a bug" button in the dashboard and a Discord. I'd love to hear which models and routing rules you'd want next.