AI API docs

One base URL, one key, 94 models. This endpoint is a drop-in replacement for the OpenAI Chat Completions API: change the base URL and the key, keep your code.

Top up & create an API key My balance & keys

1. Endpoint

Base URL	`https://memoryb2b.com/v1`
Chat completions	`POST https://memoryb2b.com/v1/chat/completions`
Auth	`Authorization: Bearer <your key>`
Models	Our display names (e.g. `DeepSeek V4.1 Flash`) or upstream ids — full list on the model discount page

Create keys in your token portal (opened with the email link). Up to 10 active keys per account; each key can carry its own spend cap.

2. First call

curl https://memoryb2b.com/v1/chat/completions \
  -H "Authorization: Bearer $MB2B_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "DeepSeek V4.1 Flash",
       "messages": [{"role": "user", "content": "One-line summary please."}],
       "stream": false}'

With the official OpenAI SDKs only two lines change — base_url and api_key:

from openai import OpenAI
client = OpenAI(base_url="https://memoryb2b.com/v1", api_key=os.environ["MB2B_KEY"])
r = client.chat.completions.create(
    model="DeepSeek V4.1 Flash",
    messages=[{"role": "user", "content": "hi"}])
print(r.choices[0].message.content)

3. Streaming

Pass "stream": true for Server-Sent Events, OpenAI format (text/event-stream, final usage chunk included). Streaming is the default when you omit the field.

4. Billing

Prepaid USD balance; charged per token at the model's discounted rate.
Your discount tier follows your lifetime top-up — see the tier table (up to 40% off).
Responses carry usage; your portal shows spend per key and per model.

5. Errors

`401`	Missing / disabled key
`402`	Balance is $0 — top up at /tokens/
`404`	Unknown model name
`502 / 503`	Upstream supplier issue — retry or pick another model

Questions about a model, a price or volume terms: email aya@memoryb2b.com or start the on-page chat.