The published model IDs and familiar JSON shapes. Feature-specific support is documented only after production verification, so those familiar identifiers do not imply a complete compatibility matrix.

Which models are live?

Nine token models: GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-5.5, GPT-5.4, GPT-5.4 Mini, Claude Opus 4.7, Claude Sonnet 4.6, and Claude Haiku 4.5. Each model card lists its current model ID and price.

How does billing work?

Billing is pay-as-you-go through prepaid credits, with no subscription tier and a $5 minimum top-up. The pricing page shows vendor-published rates beside corouter rates.

Do you support Anthropic prompt caching and OpenAI cached input?

The published OpenAI cached-input rate and Anthropic cache-read multiplier are honored 1:1 in the rate card. The pricing page lists the resulting cached-input prices where available.

Is corouter affiliated with the model vendors?

No. corouter is not affiliated with OpenAI or Anthropic. Vendor names identify the models and published prices in the catalog.

Founding-developer pricing — live now

AI tokens at outlet prices.

GPT and Claude — billed at 20% of list. Same model IDs. Familiar JSON shapes. Pay-as-you-go credits.

Get API key See pricing

Official list price

GPT-5.5

Input

$5.00/M

Output

$30.00/M

1.1M context

vs↓

On corouter

GPT-5.5

Input

$1.00/M

Output

$6.00/M

1.1M context

−80%

5× cheaper · same model ID

Outlet pricing

Every model, in plain dollars.

What providers publish, with a line through it. What you pay, in bold. Same model IDs, familiar JSON shapes, and cache discounts honored 1:1.

GPT-5.6 Sol

gpt-5.6-sol

−80%

5× cheaper

Input / Mtok: $5.00; $1.00
Output / Mtok: $30.00; $6.00

1.1M contextcache read · $0.100

GPT-5.6 Terra

gpt-5.6-terra

−80%

5× cheaper

Input / Mtok: $2.50; $0.500
Output / Mtok: $15.00; $3.00

1.1M contextcache read · $0.050

GPT-5.6 Luna

gpt-5.6-luna

−80%

5× cheaper

Input / Mtok: $1.00; $0.200
Output / Mtok: $6.00; $1.20

1.1M contextcache read · $0.020

GPT-5.5

gpt-5.5

−80%

5× cheaper

Input / Mtok: $5.00; $1.00
Output / Mtok: $30.00; $6.00

1.1M contextcache read · $0.100

GPT-5.4

gpt-5.4

−80%

5× cheaper

Input / Mtok: $2.50; $0.500
Output / Mtok: $15.00; $3.00

1M contextcache read · $0.050

GPT-5.4 Mini

gpt-5.4-mini

−80%

5× cheaper

Input / Mtok: $0.750; $0.150
Output / Mtok: $4.50; $0.900

400K contextcache read · $0.015

Claude Haiku 4.5

claude-haiku-4-5

−80%

5× cheaper

Input / Mtok: $1.00; $0.200
Output / Mtok: $5.00; $1.00

200K contextcache read · $0.020

Claude Sonnet 4.6

claude-sonnet-4-6

−80%

5× cheaper

Input / Mtok: $3.00; $0.600
Output / Mtok: $15.00; $3.00

1M contextcache read · $0.060

Claude Opus 4.7

claude-opus-4-7

−80%

5× cheaper

Input / Mtok: $5.00; $1.00
Output / Mtok: $25.00; $5.00

1M contextcache read · $0.100

See full price table Browse models

Explore

Start with the facts you need.

Compare published prices, inspect the nine-model catalog, estimate a monthly token bill, or read how the arithmetic works.

Compare prices

See vendor-published input and output rates beside corouter prices.

Open pricing

Browse live models

Review model IDs, context windows, modalities, and current availability.

Open models

Estimate API cost

Model monthly input, cached input, and output token costs before you buy.

Open calculator

Read the cost guides

How prompt caching, model tiers, and vendor list prices move a monthly bill.

Open blog

Of official list price

0×

Cheaper than retail

Dollar minimum to start

Live token models

Why corouter

Three facts you can verify before buying.

Published prices, side by side.

Each live model shows its vendor-published input and output rates beside the corouter price.

Familiar request shapes.

Use the same model IDs and familiar JSON shapes. Feature-specific compatibility is stated only after production verification.

Pay only for usage.

Prepaid credits have no subscription tier. The minimum top-up is $5, and the public rate card applies to usage.

Catalog

Nine GPT and Claude token models live now.

GPT-5.6 Sol

$30.00 $6.00 out / Mtok

GPT-5.6 Terra

$15.00 $3.00 out / Mtok

GPT-5.6 Luna

$6.00 $1.20 out / Mtok

GPT-5.5

$30.00 $6.00 out / Mtok

GPT-5.4

$15.00 $3.00 out / Mtok

GPT-5.4 Mini

$4.50 $0.900 out / Mtok

Claude Haiku 4.5

$5.00 $1.00 out / Mtok

Claude Sonnet 4.6

$15.00 $3.00 out / Mtok

Claude Opus 4.7

$25.00 $5.00 out / Mtok

GPT-5.6 Sol

$30.00 $6.00 out / Mtok

GPT-5.6 Terra

$15.00 $3.00 out / Mtok

GPT-5.6 Luna

$6.00 $1.20 out / Mtok

GPT-5.5

$30.00 $6.00 out / Mtok

GPT-5.4

$15.00 $3.00 out / Mtok

GPT-5.4 Mini

$4.50 $0.900 out / Mtok

Claude Haiku 4.5

$5.00 $1.00 out / Mtok

Claude Sonnet 4.6

$15.00 $3.00 out / Mtok

Claude Opus 4.7

$25.00 $5.00 out / Mtok

GPT-5.6 Sol

$30.00 $6.00 out / Mtok

GPT-5.6 Terra

$15.00 $3.00 out / Mtok

GPT-5.6 Luna

$6.00 $1.20 out / Mtok

GPT-5.5

$30.00 $6.00 out / Mtok

GPT-5.4

$15.00 $3.00 out / Mtok

GPT-5.4 Mini

$4.50 $0.900 out / Mtok

Claude Haiku 4.5

$5.00 $1.00 out / Mtok

Claude Sonnet 4.6

$15.00 $3.00 out / Mtok

Claude Opus 4.7

$25.00 $5.00 out / Mtok

GPT-5.6 Sol

$30.00 $6.00 out / Mtok

GPT-5.6 Terra

$15.00 $3.00 out / Mtok

GPT-5.6 Luna

$6.00 $1.20 out / Mtok

GPT-5.5

$30.00 $6.00 out / Mtok

GPT-5.4

$15.00 $3.00 out / Mtok

GPT-5.4 Mini

$4.50 $0.900 out / Mtok

Claude Haiku 4.5

$5.00 $1.00 out / Mtok

Claude Sonnet 4.6

$15.00 $3.00 out / Mtok

Claude Opus 4.7

$25.00 $5.00 out / Mtok

FAQ

The questions every dev asks first.

Pay as you go

Compare the numbers, then choose.

Nine live token models, published model IDs, familiar JSON shapes, and a $5 minimum top-up.

Get API key See pricing

AI tokens at outlet prices.

Every model, in plain dollars.

Start with the facts you need.

Compare prices

Browse live models

Estimate API cost

Read the cost guides

Three facts you can verify before buying.

Published prices, side by side.

Familiar request shapes.

Pay only for usage.

Nine GPT and Claude token models live now.

The questions every dev asks first.

What stays familiar?

Which models are live?

How does billing work?

Do you support Anthropic prompt caching and OpenAI cached input?

Is corouter affiliated with the model vendors?

Compare the numbers, then choose.