nova-api/api/config/config.yml

start-credits: 1000 # Credits given to new users.

# Credit cost per API request.
costs:
  other: 5 # Other endpoints

  chat-models: # chat completions
    gpt-4-32k: 100
    gpt-4: 20
    gpt-3: 5

## Roles Explanation

# Bonuses: They are a multiplier for costs
#          They work like: final_cost = cost * bonus

roles:
  owner:
    bonus: 0
  enterprise:
    bonus: 0.2
  admin:
    bonus: 0.4
  helper:
    bonus: 0.6
  default:
    bonus: 1.0
Documented almost everything 2023-10-12 00:03:15 +02:00			`start-credits: 1000 # Credits given to new users.`
Re-added rate limiting, fixed a bunch of issues 2023-08-17 13:11:35 +02:00
Documented almost everything 2023-10-12 00:03:15 +02:00			`# Credit cost per API request.`
Re-added rate limiting, fixed a bunch of issues 2023-08-17 13:11:35 +02:00			`costs:`
Documented almost everything 2023-10-12 00:03:15 +02:00			`other: 5 # Other endpoints`
Re-added rate limiting, fixed a bunch of issues 2023-08-17 13:11:35 +02:00
Documented almost everything 2023-10-12 00:03:15 +02:00			`chat-models: # chat completions`
Added token-based credit cost 2023-10-16 23:34:54 +02:00			`gpt-4-32k: 100`
			`gpt-4: 20`
			`gpt-3: 5`
Re-added rate limiting, fixed a bunch of issues 2023-08-17 13:11:35 +02:00
			`## Roles Explanation`

			`# Bonuses: They are a multiplier for costs`
			`# They work like: final_cost = cost * bonus`

			`roles:`
			`owner:`
Special: halved GPT-3 credits cost (thanks for 1.000 members!) If no model is given, the API now defaults to gpt-3.5-turbo We now also support gpt-3.5-turbo-0301! Made provider code asynchronous New dependency requirement: aiofiles Staff now gets a notification when a provider key is invalid Internal improvements with log webhooks for staff Removed image model check 2023-09-02 21:15:55 +02:00			`bonus: 0`
Added token-based credit cost 2023-10-16 23:34:54 +02:00			`enterprise:`
Special: halved GPT-3 credits cost (thanks for 1.000 members!) If no model is given, the API now defaults to gpt-3.5-turbo We now also support gpt-3.5-turbo-0301! Made provider code asynchronous New dependency requirement: aiofiles Staff now gets a notification when a provider key is invalid Internal improvements with log webhooks for staff Removed image model check 2023-09-02 21:15:55 +02:00			`bonus: 0.2`
Added token-based credit cost 2023-10-16 23:34:54 +02:00			`admin:`
Re-added rate limiting, fixed a bunch of issues 2023-08-17 13:11:35 +02:00			`bonus: 0.4`
Added token-based credit cost 2023-10-16 23:34:54 +02:00			`helper:`
Special: halved GPT-3 credits cost (thanks for 1.000 members!) If no model is given, the API now defaults to gpt-3.5-turbo We now also support gpt-3.5-turbo-0301! Made provider code asynchronous New dependency requirement: aiofiles Staff now gets a notification when a provider key is invalid Internal improvements with log webhooks for staff Removed image model check 2023-09-02 21:15:55 +02:00			`bonus: 0.6`
Re-added rate limiting, fixed a bunch of issues 2023-08-17 13:11:35 +02:00			`default:`
I don't even fkin know lmfao 2023-08-18 21:23:00 +02:00			`bonus: 1.0`