inferroute

Pay only for what you use.

No subscription. The open frontier — Kimi, GLM, MiniMax — sits well under market on every bucket: input, cache, and output. Claude (Anthropic) is pass-through on your own key, never marked up.

Open models, per token

vs OpenRouter

GLM 5.2

Economy
$0.19/ M effective

−48% vs OpenRouter $0.37/M

Economy · per token
Input$0.49 $0.95−48%Cached$0.099 $0.18−45%Output$1.40 $3.00−53%
On-demand
Input$0.90Cached$0.19Output$3.40

Kimi K2.7

Economy
$0.12/ M effective

−61% vs OpenRouter $0.31/M

Economy · per token
Input$0.29 $0.74−61%Cached$0.059 $0.15−61%Output$1.40 $3.50−60%
On-demand
Input$0.54Cached$0.090Output$2.40

MiniMax M3

Economy
$0.039/ M effective

−68% vs OpenRouter $0.12/M

Economy · per token
Input$0.090 $0.30−70%Cached$0.019 $0.060−68%Output$0.44 $1.20−63%
On-demand
Input$0.24Cached$0.049Output$0.84

GLM 5.1

Economy
$0.13/ M effective

−79% vs OpenRouter $0.62/M

Economy · per token
Input$0.34 $0.98−65%Cached$0.069 $0.49−86%Output$0.90 $3.08−71%
On-demand
Input$0.69Cached$0.14Output$2.40

Kimi K2.6

Economy
$0.11/ M effective

−64% vs OpenRouter $0.29/M

Economy · per token
Input$0.24 $0.66−64%Cached$0.049 $0.14−66%Output$1.40 $3.41−59%
On-demand
Input$0.49Cached$0.099Output$2.40

Kimi K2.5

Economy
$0.061/ M effective

−62% vs OpenRouter $0.16/M

Economy · per token
Input$0.14 $0.38−63%Cached$0.029 $0.075*−61%Output$0.74 $2.02−63%
On-demand
Input$0.34Cached$0.069Output$1.40

MiniMax M2.7

Economy
$0.036/ M effective

−63% vs OpenRouter $0.098/M

Economy · per token
Input$0.084 $0.24−65%Cached$0.019 $0.048*−60%Output$0.34 $0.96−65%
On-demand
Input$0.19Cached$0.039Output$0.69

MiniMax M2.5

Economy
$0.024/ M effective

−71% vs OpenRouter $0.082/M

Economy · per token
Input$0.054 $0.15−64%Cached$0.009 $0.050−82%Output$0.44 $0.90−51%
On-demand
Input$0.090Cached$0.019Output$0.84

The big figure is the effective $/M — the three buckets blended on real agent traffic (19.4% input / 79.2% cached / 1.4% output). Per-token rates below in $/M. Economy is the deferred, best-effort lane (ideal for agents that can wait); on-demand is real-time. Strikethrough is OpenRouter — use the arrows, dots, or swipe to browse models. * imputed cache rate where the provider doesn't publish one.

Works with the Claude you already pay for. Sonnet & Opus pass through your own Anthropic subscription — never marked up, never touched. InferRoute just adds the cheaper open models on top.

Two ways to pay

Most people

One account, no subscription

Top up a balance and pay only for what you route — no monthly bill, stop anytime.

  • No usage-cap anxiety — never ration prompts or get throttled
  • Prepaid credits that never expire
  • Top up by card or USDC — 0% fees in USDC
Get started
x402 enabled

Pay-as-you-go in USDC

Approve once via the open x402 standard. Usage is metered per request and topped up in $1 USDC increments — batched on purpose, so you pay 0 network fees.

  • $1 gasless top-ups, drawn down per request — 0% fees
  • Non-custodial: you approve, funds stay yours until used
  • Built for agents and machine-to-machine spend
Learn about x402

Pay only for what you use. No subscription, no commitment.