⚡ Routing Engine · Layer L251

Route any AI task to the cheapest capable model — automatically.

Selects the optimal model for every task — cheapest for simple tasks, most capable for complex ones. One key, all providers, automatic selection. 78p back on every £1.

Start building free View docs Pricing

Live £29/mo or £290/yr — save 17%

78%Earnings back

< 50msRouting latency

6+Model providers

FreeTier available

Auto model selectionCost optimisedWORM-sealedAll providers78% backFree tier

How it works

Routing Engine in three steps

Production-grade and live on api.forcedream.ai. One Bearer token, zero extra setup.

01 🔍
Classify

The engine classifies your task as simple/medium/complex/critical. You can set quality floors and cost ceilings per call.
POST /v1/route
02 ⚡
Route

The cheapest model that meets your quality threshold is selected. Automatic failover if the primary model errors.
model_used: claude-haiku
03 💚
Earn

78% of every routed call's revenue flows back to you. Earnings accumulate in your wallet and withdraw via Stripe.
earned_pence: 0.78

What's included

Everything you need, nothing you don't

✓Automatic model selection — simple/medium/complex/critical

✓Budget constraints — cap spend per call in pence

✓Latency constraints — cap response time in milliseconds

✓Model fallback — automatic failover on provider error

✓WORM-sealed routing records — auditable model selection

✓78% of earnings back on every routed call

Quick start

api.forcedream.ai POST /v1/route

FORCEDREAM OS · L251 LIVE

$ curl https://api.forcedream.ai/v1/route \
  -H "Authorization: Bearer $KEY" \
  -d '{"task":"summarise this document","input":"Your content here"}'

→ {"output":"...","model_used":"claude-haiku","cost_pence":0.04,"earned_pence":0.78}

# With quality/cost constraints
$ curl https://api.forcedream.ai/v1/route \
  -H "Authorization: Bearer $KEY" \
  -d '{"task":"draft a legal contract","input":"...","quality":"high","max_cost_pence":200}'

→ {"output":"...","model_used":"claude-opus","cost_pence":180,"earned_pence":140.4}

Pricing

Simple, transparent pricing

78% of API earnings flow back to you on every call. No hidden fees. Free tier available.

Starter

£29

/mo · £290/yr

1M routed calls/month

All model providers

Auto model selection

78% earnings

Standard failover

Start free →

How Routing Engine compares

Purpose-built for AI products. Not retrofitted from general-purpose tools.

Feature	ForceDream Routing	LangChain Router	OpenRouter	DIY
Auto model selection	✓	Manual config	Limited	Manual
Cost optimisation	✓	Manual	—	Manual
80% earnings back	✓	—	—	—
WORM audit	✓	—	—	—
Failover on error	✓	Manual	Partial	Manual
Free tier	✓	✓	✓	✓

FAQ

Frequently asked questions

How does model selection work?

The engine classifies your task as simple/medium/complex/critical and picks the cheapest model that meets the quality threshold. Simple tasks use fast cheap models. Critical tasks use the most capable.

Which models are available?

Claude (Haiku/Sonnet/Opus), Gemini (Flash/Pro), GPT-4o and more. The engine selects across providers automatically. You can whitelist or blacklist specific models per API key.

Can I force a specific model?

Yes — pass model:"claude-sonnet-4" in the request body to pin the model. The engine still logs, earns and audits the call.

What does max_cost_pence do?

It guarantees the call costs under the specified pence amount, falling back to a cheaper model if needed. If no model meets the budget, the call returns an error rather than overspending.

How do I earn 78p per £1?

Every call through /v1/route earns you 78% of gross API revenue. Earnings accumulate in your wallet and are withdrawable via Stripe at any time.