hyphen

Flat-rate AI gateway

Frontier coding agents, without frontier prices.

Point Claude Code, Codex, or any OpenAI-compatible agent at one URL. Every frontier model, one flat monthly price, a hard token cap — roughly a tenth of the bill.

bash
# Point any OpenAI-compatible agent at Hyphen
export OPENAI_BASE_URL="https://api.hyphen-solution.com/v1"
export OPENAI_API_KEY="hyp_sk_your_key"

# Your agent just works — no SDK changes
curl "$OPENAI_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{"model":"m3","messages":[{"role":"user","content":"ship it"}]}'

Drop-in for Claude Code · Codex CLI · Cursor · Cline · opencode · any OpenAI-compatible client

The same class of work. About a tenth of the rate.

Frontier providers meter every token, and the bill scales with your agent. Hyphen routes to high-capability models at a fraction of the per-token cost — then charges one flat monthly price.

Buying frontier APIs directly

Input · per 1M tokens
$3.00
Output · per 1M tokens
$15.00
Billing
Metered · scales with use

On Hyphen

Input · per 1M tokens
$0.30
Output · per 1M tokens
$1.20
Billing
Flat · from $3/mo, hard cap

Frontier rates shown are published list prices for a leading frontier coding model. You never pay per token on Hyphen — plans are flat, with a token allowance you can see.

One key. A catalog that keeps growing.

Hyphen is a multi-model gateway, not a single-provider reseller. Reach every model by capability through the same endpoint — pick by the job, not the vendor. The roster only grows.

m3Flagship

Deepest reasoning. Reach for it on hard, multi-step agent work and long refactors.

200K context

m2.7Balanced

Fast, strong all-rounder — the daily driver for most coding and agent tasks.

205K context

m2.5Efficient

Quickest and cheapest. Great for high-volume, lighter calls and tool loops.

200K context

Coming soonDeepSeekGLMKimi+ more

Running in about a minute.

01

Create a key

Sign up and generate an API key in the console. Set a monthly token allowance you control.

02

Set one base URL

Point your agent at the gateway with two environment variables. Nothing else changes.

OPENAI_BASE_URL=https://api.hyphen-solution.com/v1
OPENAI_API_KEY=hyp_sk_…
03

Run your agent

Claude Code, Codex, Cursor — unchanged. Watch tokens draw down live in the console.

One flat price. A token cap you can see.

Pick an allowance. Hit it and requests pause with a clean 429 until the next cycle — never a surprise invoice. Need more mid-month? Top up with credit that never expires.

Starter

Just try it.
3Mtokens / month

$3 per month

  • ~3M tokens / month
  • Every model, one endpoint
  • Bring any OpenAI-compatible agent
  • Community support
Start for $3
Most popular

Pro

The daily driver.
13Mtokens / month

$12 per month

  • ~13M tokens / month
  • Every model, one endpoint
  • Priority routing
  • Add prepaid credit anytime
  • Email support
Get started

Team

Teams & agent fleets.
33Mtokens / month

$29 per month

  • ~33M tokens / month
  • Everything in Pro
  • Highest-priority routing
  • Priority support
Get started

See full pricing and credit packs

Keep your agent. Cut the bill.

Create a key, set your cap, and point your agent at Hyphen. Start for $3 — it takes about a minute.