Base URL
https://100milionseek.top/v1
API REFERENCE
One model, one OpenAI-compatible endpoint, one daily allowance. Everything you need is on this page.
Base URL
https://100milionseek.top/v1
Authentication
Authorization: Bearer hms-…
Model
deepseek-v4.1-flash-free
198K tokens · 64K tokens
curl https://100milionseek.top/v1/chat/completions \
-H "Authorization: Bearer $HMS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v4.1-flash-free",
"stream": true,
"messages": [{"role": "user", "content": "Explain SQLite WAL mode in 3 lines"}]
}'import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["HMS_API_KEY"],
base_url="https://100milionseek.top/v1",
)
stream = client.chat.completions.create(
model="deepseek-v4.1-flash-free",
messages=[{"role": "user", "content": "Explain SQLite WAL mode in 3 lines"}],
stream=True,
)
for chunk in stream:
if chunk.choices:
print(chunk.choices[0].delta.content or "", end="")import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.HMS_API_KEY,
baseURL: "https://100milionseek.top/v1",
});
const { data, response } = await client.chat.completions
.create({
model: "deepseek-v4.1-flash-free",
messages: [{ role: "user", content: "Summarise this changelog" }],
})
.withResponse();
console.log(data.choices[0].message.content);
console.log("tokens left today:", response.headers.get("x-quota-remaining"));
Each account gets 100,000,000 tokens per UTC day. A finished request adds prompt_tokens + completion_tokens from the model's usage report. A stream you cancel is counted from the prompt size and the text already sent. The counter resets at 00:00 UTC.
All 100,000,000 tokens of each UTC day stream at full speed, and the header x-qos always reads full.
Every API response carries these headers:
x-quota-limit: 100000000 x-quota-used: 18342 x-quota-remaining: 99981658 x-quota-reset: 2026-11-13T00:00:00.000Z
When the allowance is used up, the API answers like this until the reset:
HTTP/1.1 429 Too Many Requests
Retry-After: 30512
x-quota-remaining: 0
{"error":{"message":"You have used today's 100,000,000 tokens. The allowance resets at 2026-11-13T00:00:00.000Z (00:00 UTC).",
"type":"rate_limit_error","code":"daily_quota_exceeded",
"limit":100000000,"used":100000412,"reset_at":"2026-11-13T00:00:00.000Z","retry_after_s":30512}}
| Route | What it does | Notes |
|---|---|---|
GET /v1/models | Lists the one model. | Needs a key. |
POST /v1/chat/completions | Chat completion, streaming or not. | Needs a key. Counts against your daily allowance. |
POST /v1/systemone | JEV judge call: model, state and questions in, one probability per question out. | Needs a key. Each token counts as 0.2 against your daily allowance. |
GET /api/status | Live service numbers as JSON. | No key needed. |
GET /health | Liveness check. | No key needed. |
Up to 5 requests per account can run at the same time. One more gets 429 session_limit_reached; wait for one to finish.
| Status and code | Meaning | What to do |
|---|---|---|
401 missing_api_key / invalid_api_key | The key is missing or wrong. | Send Authorization: Bearer <key>. Sign in to rotate the key if you lost it. |
400 invalid_json / invalid_messages | The request body is not valid. | Check the JSON and the messages array. |
400 context_length_exceeded | The input is longer than the context window. | Shorten the conversation or drop old messages. |
404 model_not_found | Unknown model. | Use deepseek-v4.1-flash-free. |
413 body_too_large | The body is larger than 4 MB. | Send less text per request. |
429 daily_quota_exceeded | Today's allowance is used up. | Wait until the time in reset_at (00:00 UTC). |
429 session_limit_reached | Too many requests from your account at once. | Wait for a running request to finish. |
503 upstream_saturated / upstream_busy | The service is at full capacity. | Retry after the Retry-After seconds. |
502 upstream_error | The model failed to answer. | Retry. Failed requests do not count against your allowance. |
Announcement
Thanks to the amazing response, we have removed our speed limit! Enjoy 100 million tokens to your heart's content!!
Incident report · Nov 15, 2026
On Nov 15, 2026, from 06:20 to 07:52 UTC, the service slowed down and many requests failed because a database lookup overloaded our server. We fixed it at 07:52 UTC and the service has run normally since.
10 TOKEN RESET vouchers for your account
Every account that existed at the time gets 10 vouchers. One voucher sets the tokens you have used today back to zero, so your full daily allowance is available again.
Use one with the button on your dashboard, or call the API with your key:
curl -X POST https://100milionseek.top/api/vouchers/redeem \
-H "Authorization: Bearer $HMS_API_KEY"