Pick a MiMo-V2.6 edition
MiMo-V2.6 Flash
1M tokens context · up to 128K tokens out
$0.189 per 1M input · $0.378 per 1M output
xiaomi/mimo-v2.6-flash
MiMo-V2.6 Pro
1M tokens context · up to 128K tokens out
$0.58725 per 1M input · $1.1745 per 1M output
xiaomi/mimo-v2.6-pro
MiMo-V2.6 Pro UltraSpeed
1M tokens context · up to 128K tokens out
$5.8725 per 1M input · $11.745 per 1M output
xiaomi/mimo-v2.6-pro-ultraspeed
Make your first MiMo-V2.6 call
# MiMo-V2.6 through the OpenAI-compatible chat endpoint
curl https://apimany.com/v1/chat/completions \
-H "Authorization: Bearer $APIMANY_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"xiaomi/mimo-v2.6-pro","messages":[{"role":"user","content":"Summarise this contract in five bullets."}],"max_tokens":512}'GET /v1/models are callable. Catalog entries are not availability promises. Why route MiMo-V2.6 through API Many
Adding one more model family normally means one more credential, one more billing relationship and one more adapter in every service that needs it. API Many keeps the public model id stable and separate from the route behind it, so MiMo-V2.6 arrives as three new ids on the endpoint you already call.
Requests are quoted and reserved against your USD balance before dispatch, then settled from the token usage the model reports. Every attempt is recorded, and an uncertain outcome is held for reconciliation instead of being silently retried — which matters most on the long-context prompts MiMo-V2.6 is built for.
Choosing between the editions
Start on Flash for classification, extraction and other high-volume work where the per-token rate dominates. Move to Pro when answer quality matters more than unit cost. Use Pro UltraSpeed when you need Pro-level output inside an interactive latency budget; it is the same flagship checkpoint tuned for throughput, and it is priced accordingly.
MiMo-V2.6 API questions
How do I call the MiMo-V2.6 API?+
Point any OpenAI SDK or plain HTTP client at POST /v1/chat/completions with an API Many key, send xiaomi/mimo-v2.6-pro (or the Flash or Pro UltraSpeed id) as the model, and pass a normal messages array. No MiMo-specific client is required.
Which MiMo-V2.6 editions are in the catalog?+
Three: MiMo-V2.6 Flash for high-volume work, MiMo-V2.6 Pro as the flagship, and MiMo-V2.6 Pro UltraSpeed for the same Pro quality at much lower latency. All three publish a 1M-token context window.
What does the MiMo-V2.6 API cost?+
Per token, in USD. Flash starts at $0.189 per million input tokens; each edition's exact input and output rate is published on its model page and by authenticated GET /v1/models. Token usage comes back on every response so you can reconcile the bill.
Is the MiMo-V2.6 API OpenAI-compatible?+
Yes for non-streaming text chat: the same request body, the same response shape and the same usage object. Streaming, audio, embeddings and multimodal chat input are outside the current public contract.
Is MiMo-V2.6 callable on my account right now?+
Catalog entries are not availability promises. Authenticated GET /v1/models is the single source of truth: if xiaomi/mimo-v2.6-pro is listed there with a rate, it is enabled for your key.