AnyInt Intelligent Routing

Intelligent routing: make every model call earn its cost

Plain routing sends every request to one fixed model. Intelligent routing first understands the request, then picks the best-fit model and endpoint under quality, cost, latency, privacy, and safety constraints.

User requestExplain this invoice line item and tell me whether a human agent needs to review it.
Signal BrainLow-risk, repetitive business question with a warm cache.
Candidate model pools

Billing support questionLightweight instruct models

Multi-file code refactorFlagship coding models

Patient record summaryAttested private deployments

Long document synthesisLong-context flagship models

Routing Arena

Same goal, lower cost with intelligent routing

The numbers below are local demo data used to illustrate the product advantage: without lowering the quality threshold, simple requests go to cheaper models while high-risk or hard tasks stay on dedicated lanes.

Lightweight instruct models

Billing support question

82%Cost index
41%Latency
92%Quality confidence
-82%

Versus routing everything to a fixed strong model, it keeps the quality threshold while cutting unnecessary large-model calls.

Signal panel

DomainSupport / Billing
ComplexitySimple
Context size~2K tokens
Sensitive spansNone detected
Cache similarity0.91
Normal Route vs Smart Route

Fixed routing answers “can it run”; smart routing answers “is it worth it”

How plain routing runs

Every request goes to the same model or a set of static weights, so even simple tasks burn an expensive model.

01Input requests

Support, code, privacy, and long-document requests are all treated the same, with no differentiation.

02Fixed strong model

All traffic goes through one lane — no signal analysis and no policy branching.

03Uniform output

Quality is acceptable, but even the simplest tasks burn the most expensive compute.

How intelligent routing runs

Keep one complete routing chain: pick a task type on the left, match its rule and policy in the middle, and land on different model pools and endpoints on the right.

011. Signal extraction

Classify task type, complexity, context, privacy, and cache similarity.

  • Lightweight tasks
  • General tasks
  • Strong reasoning tasks
  • Private tasks
022. Matching & policy

Merge signals into a decision vector, then hit concrete rules by priority.

complexity = simple AND cache_similarity > 0.9
033. Plugin chain

Run semantic cache, cost gates, prompt injection, verification, and redaction.

  • Semantic cache lookup
  • Lightweight model pool
  • Support knowledge injection
044. Model & endpoint

Select the candidate pool by quality, cost, latency, and deployment boundaries.

Target model pool
Lightweight instruct models
Serving endpoint
AnyInt low-cost pool
Average cost-18%
Plain route
Smart route
Low-risk request latency-21%
Plain route
Smart route
Quality pass rate+4%
Plain route
Smart route
Sensitive-data interception+38%
Plain route
Smart route

Signals in, policy out

Routing is a configurable orchestration layer, not random dispatch. Multi-dimensional signals profile every request, rules compose them into policies, and each policy binds the plugin chain that runs before, during, and after the model call.

Cost policies

When complexity and cache signals allow, check the semantic cache first, then climb from lightweight pools only as far as quality demands.

Quality policies

When a fact-check or schema signal fires, add verification and repair stages instead of blindly paying for a bigger model.

Compliance policies

When sensitive spans are detected, restrict the candidate pool to attested deployments, redact before logging, and write an audit record.