One key.
One flat price.
Real models under their real names, for your coding agent.
Cursor
Cline
opencodeAiderZedGooseOpenHandsKilo CodeHermes agentOpenClaw
LangChain
n8nVercel AI SDK
$4/mo
Cancel anytime.
10models
Real names.
2env vars
The whole setup.
$0overage
Ever.
An honest middleman.
The response tells you exactly which model answered.
Real names
minimax-m3 is MiniMax's M3. Nothing is rebranded.
Hard cap
At the cap, requests pause with a 429 until next cycle. Overage billing does not exist here.
Transparent serving
Every response reports the model that served it, including during a fallback.
A catalog that keeps growing.
Ten MiniMax models today. The daily drivers:
MiniMax's strongest. Use it for hard multi-step agent work, long refactors, and deep reasoning.
~205K contextby MiniMax
The daily driver. Fast, strong all-rounder for most coding and agent tasks.
~205K contextby MiniMax
Quickest of the family. Great for high-volume, lighter calls and tool loops.
~205K contextby MiniMax
Long-context text model. Feed it whole codebases or big document sets in one prompt.
Long contextby MiniMax
Running in about a minute.
Create a key
Pick a plan or a credit pack. Your key appears in the console.
Set one base URL
Two environment variables. Nothing else changes.
BASE_URL=https://api.hyphen-solution.com/v1 API_KEY=sk-…
Run your agent
Codex, Cursor, and opencode run unchanged.
Where Hyphen actually is.
This is new. It launched in July 2026, it is built and run by one developer, and the catalog is one provider today with more being added. No funding, no waitlist, no enterprise tier.
What that buys you: the rate card is public, every response names the model that served it, and your spend is capped so a runaway agent costs you nothing beyond the plan price. If something breaks, email support@hyphen-solution.com and a person reads it.
What it does not buy you: an uptime guarantee, a support SLA, or a track record. Start on the $4 plan and see whether it holds up before you trust it with anything that matters.
A flat price and a hard monthly cap.
Starter
Just try it.~3.8M tokens a month, hard cap.
- 1 API key
- Monthly allowance with a hard cap
- Full model catalog, one endpoint
- Email support
Pro
The daily driver.~19M tokens a month, hard cap.
- 3 API keys
- A bigger monthly allowance for daily agent use
- Email support, prioritized
Team
For teams and agent fleets.~46M tokens a month, hard cap.
- Unlimited API keys
- The largest monthly allowance, pooled across all your keys
- A key per teammate, all drawing on that one allowance
- Email support