Create an endpoint

POST

provider

endpoints

cURL

curl --request POST \
  --url https://api.tokenchange.io/v1/provider/endpoints \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model_id": "<string>",
  "pricing": {
    "cached_per_million": "<string>",
    "input_per_million": "<string>",
    "output_per_million": "<string>"
  },
  "slots": 123,
  "url": "<string>",
  "bearer_token": "<string>",
  "capabilities": [
    "<string>"
  ],
  "context_window": 123,
  "quantization": "<string>",
  "upstream_model_name": "<string>"
}
'

import requests

url = "https://api.tokenchange.io/v1/provider/endpoints"

payload = {
    "model_id": "<string>",
    "pricing": {
        "cached_per_million": "<string>",
        "input_per_million": "<string>",
        "output_per_million": "<string>"
    },
    "slots": 123,
    "url": "<string>",
    "bearer_token": "<string>",
    "capabilities": ["<string>"],
    "context_window": 123,
    "quantization": "<string>",
    "upstream_model_name": "<string>"
}
headers = {
    "Authorization": "Bearer <token>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    model_id: '<string>',
    pricing: {
      cached_per_million: '<string>',
      input_per_million: '<string>',
      output_per_million: '<string>'
    },
    slots: 123,
    url: '<string>',
    bearer_token: '<string>',
    capabilities: ['<string>'],
    context_window: 123,
    quantization: '<string>',
    upstream_model_name: '<string>'
  })
};

fetch('https://api.tokenchange.io/v1/provider/endpoints', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

{
  "availability": "pending",
  "capabilities": [
    "<string>"
  ],
  "config_version": 123,
  "created_at": "<string>",
  "id": "<string>",
  "latest_validations": [
    {
      "config_version": 123,
      "finished_at": "<string>",
      "id": "<string>",
      "metrics": {},
      "started_at": "<string>",
      "status": "passed",
      "suite": "<string>",
      "trigger": "admission",
      "error": "<string>"
    }
  ],
  "model_id": "<string>",
  "pricing": {
    "input_per_million": "<string>",
    "output_per_million": "<string>",
    "cached_per_million": "<string>"
  },
  "slots": 123,
  "slug": "<string>",
  "status": "active",
  "url": "<string>",
  "context_window": 123,
  "latest_price_change": {
    "id": "<string>",
    "pricing": {
      "input_per_million": "<string>",
      "output_per_million": "<string>",
      "cached_per_million": "<string>"
    },
    "requested_at": "<string>",
    "status": "pending",
    "decided_at": "<string>",
    "decision_reason": "<string>"
  },
  "output_tokens_per_sec": 123,
  "quantization": "<string>",
  "review_reason": "<string>",
  "reviewed_at": "<string>",
  "upstream_model_name": "<string>",
  "validated_at": "<string>",
  "validation_error": "<string>"
}

{
  "error": {
    "code": "<string>",
    "message": "<string>",
    "type": "<string>",
    "resource_id": "<string>"
  }
}

Authorizations

Authorization

string

header

required

A gateway API token (tcg_…) created in the dashboard

Headers

Idempotency-Key

string | null

Makes retries safe for 24 hours

Body

application/json

model_id

string

required

pricing

object

required

Show child attributes

slots

integer<int32>

required

url

string

required

Public HTTPS base URL of your OpenAI-compatible server.

bearer_token

string | null

Sent upstream as Authorization: Bearer …. Write-only.

capabilities

string[]

context_window

integer<int32> | null

quantization

string | null

upstream_model_name

string | null

Model name your server expects; defaults to the catalog slug.

Response

Created

One of your endpoints. The upstream bearer token is write-only and never returned.

availability

enum<string>

required

Available options:

pending,

validating,

up,

degraded,

down,

rejected

capabilities

string[]

required

config_version

integer<int32>

required

created_at

string

required

string

required

latest_validations

object[]

required

Latest run per suite against the current configuration.

Show child attributes

model_id

string

required

pricing

object

required

Prices in USD per million tokens, as decimal strings.

Show child attributes

slots

integer<int32>

required

slug

string

required

status

enum<string>

required

Available options:

active,

paused,

suspended,

retired,

pending,

rejected

url

string

required

context_window

integer<int32> | null

latest_price_change

object | null

Price changes on a live endpoint wait for operator review; the live prices stay in effect.

Show child attributes

output_tokens_per_sec

number<double> | null

quantization

string | null

review_reason

string | null

reviewed_at

string | null

upstream_model_name

string | null

validated_at

string | null

validation_error

string | null

⌘I

cURL

curl --request POST \
  --url https://api.tokenchange.io/v1/provider/endpoints \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model_id": "<string>",
  "pricing": {
    "cached_per_million": "<string>",
    "input_per_million": "<string>",
    "output_per_million": "<string>"
  },
  "slots": 123,
  "url": "<string>",
  "bearer_token": "<string>",
  "capabilities": [
    "<string>"
  ],
  "context_window": 123,
  "quantization": "<string>",
  "upstream_model_name": "<string>"
}
'

import requests

url = "https://api.tokenchange.io/v1/provider/endpoints"

payload = {
    "model_id": "<string>",
    "pricing": {
        "cached_per_million": "<string>",
        "input_per_million": "<string>",
        "output_per_million": "<string>"
    },
    "slots": 123,
    "url": "<string>",
    "bearer_token": "<string>",
    "capabilities": ["<string>"],
    "context_window": 123,
    "quantization": "<string>",
    "upstream_model_name": "<string>"
}
headers = {
    "Authorization": "Bearer <token>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    model_id: '<string>',
    pricing: {
      cached_per_million: '<string>',
      input_per_million: '<string>',
      output_per_million: '<string>'
    },
    slots: 123,
    url: '<string>',
    bearer_token: '<string>',
    capabilities: ['<string>'],
    context_window: 123,
    quantization: '<string>',
    upstream_model_name: '<string>'
  })
};

fetch('https://api.tokenchange.io/v1/provider/endpoints', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

{
  "availability": "pending",
  "capabilities": [
    "<string>"
  ],
  "config_version": 123,
  "created_at": "<string>",
  "id": "<string>",
  "latest_validations": [
    {
      "config_version": 123,
      "finished_at": "<string>",
      "id": "<string>",
      "metrics": {},
      "started_at": "<string>",
      "status": "passed",
      "suite": "<string>",
      "trigger": "admission",
      "error": "<string>"
    }
  ],
  "model_id": "<string>",
  "pricing": {
    "input_per_million": "<string>",
    "output_per_million": "<string>",
    "cached_per_million": "<string>"
  },
  "slots": 123,
  "slug": "<string>",
  "status": "active",
  "url": "<string>",
  "context_window": 123,
  "latest_price_change": {
    "id": "<string>",
    "pricing": {
      "input_per_million": "<string>",
      "output_per_million": "<string>",
      "cached_per_million": "<string>"
    },
    "requested_at": "<string>",
    "status": "pending",
    "decided_at": "<string>",
    "decision_reason": "<string>"
  },
  "output_tokens_per_sec": 123,
  "quantization": "<string>",
  "review_reason": "<string>",
  "reviewed_at": "<string>",
  "upstream_model_name": "<string>",
  "validated_at": "<string>",
  "validation_error": "<string>"
}

{
  "error": {
    "code": "<string>",
    "message": "<string>",
    "type": "<string>",
    "resource_id": "<string>"
  }
}

Get started

Models

Inference keys

Usage

Billing

Provider

Authorizations

Headers

Body

Response