MiniMax API relay: text, speech and video through one endpoint

OpenAI-compatible access to the full MiniMax line-up with transparent USD pricing. Change base_url to https://yiduochan.com/v1 and you are done.

MiniMax-M3

Flagship text model with a 1M-token context window. $0.30 input / $1.20 output per million tokens (≤512K tier), cache hits $0.06.

MiniMax-M2.7 / M2.7-highspeed

Cost-efficient chat and coding model from $0.30 per million input tokens, with a faster highspeed variant.

MiniMax-H3 video generation

Hailuo H3 multimodal video: $0.08 per second at 768P, $0.13 per second at 2K, text-to-video and image-to-video.

Speech 2.8 text to speech

speech-2.8-hd and speech-2.8-turbo, billed per character through the OpenAI-compatible /v1/audio/speech endpoint.

Why call MiniMax through YiduoChan

OpenAI-compatible API: point base_url at https://yiduochan.com/v1 and keep using the OpenAI SDKs, LangChain, Cursor, Cherry Studio and every other tool you already have.
Anthropic Messages compatible: /v1/messages is available too, so Claude Code can run on MiniMax-M3 or MiniMax-M2.7.
Transparent USD pricing: prices are converted from MiniMax's official list prices, billed on actual usage with no minimum commitment; the pricing page shows live input, output and cache prices for every model.
Text, speech and video in one place: chat completions, TTS and video generation share one API key and one balance.
Retries across upstream keys: requests rotate over multiple MiniMax API keys and are retried when one is rate-limited or fails.
Usage you can audit: the console shows tokens, latency and cost for every request by key, model and time range.

MiniMax model prices (USD)

Model	Type	Price
MiniMax-M3	Chat, 1M context	≤512K: $0.30 in / $1.20 out / $0.06 cache read; 512K to 1M: $0.60 in / $2.40 out / $0.12 cache read (per 1M tokens)
MiniMax-M2.7	Chat	$0.30 in / $1.20 out / $0.06 cache read / $0.375 cache write
MiniMax-M2.7-highspeed	Chat, high speed	$0.60 in / $2.40 out / $0.06 cache read / $0.375 cache write
speech-2.8-hd	TTS, high fidelity	$1.00 per 10k characters
speech-2.8-turbo	TTS, fast	$0.60 per 10k characters
MiniMax-H3	Video generation	768P $0.08 per second, 2K $0.13 per second; first 5 reference images free, then $0.0600 each

Three steps to your first call

Create an account and open the console.
Create an API key on the API keys page.
Send requests to https://yiduochan.com/v1 with one of the model names above.

curl https://yiduochan.com/v1/chat/completions \
  -H "Authorization: Bearer $YIDUOCHAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "MiniMax-M2.7",
    "messages": [{"role": "user", "content": "Introduce the MiniMax models in one sentence"}]
  }'

Model guides

Integration guides

FAQ

What does a MiniMax API relay do?

It sits between you and the MiniMax open platform: one OpenAI-compatible endpoint, one key, pay-as-you-go billing per token or per second, and no need to manage MiniMax accounts or their rate limits yourself.

Which MiniMax models are available?

Text models MiniMax-M3, MiniMax-M2.7 and MiniMax-M2.7-highspeed, the speech models speech-2.8-hd and speech-2.8-turbo, and the MiniMax-H3 video model. New MiniMax releases are added as they ship.

How is pricing set?

Prices are quoted and billed in US dollars, converted from MiniMax's official list prices. The pricing page always shows the rates currently in effect.

How do I use it with the OpenAI SDK?

Set base_url to https://yiduochan.com/v1, use a key created on this site as api_key and pass a MiniMax model name such as MiniMax-M2.7. Requests and responses follow the OpenAI format.

Can I use MiniMax models in Claude Code?

Yes. Set ANTHROPIC_BASE_URL to https://yiduochan.com, ANTHROPIC_AUTH_TOKEN to your key and ANTHROPIC_MODEL to MiniMax-M2.7 or MiniMax-M3.

Is streaming supported?

Yes. Set stream to true on chat completions to receive tokens over server-sent events.