MiMo-V2.6-Pro.
MiMo-V2.6-Pro is Xiaomi's MIT-licensed mixture-of-experts model, 1.02T total and 42B active, with text, image, video and audio input and a 1M-token context for long agent runs.
8× Nvidia A100 SXM4 · $5.73/hr.
Most-aggressive quantisation we have a working recommendation for. Lower precision = less VRAM = cheaper hardware, at a small accuracy cost.
Cheapest hosted endpoints.
Variants in the MiMo family.
MiMo-V2.6-Flash
MiMo-V2.6-Flash is Xiaomi's MIT-licensed 309B mixture-of-experts model, 15B a...
MiMo-V2.6-Pro-UltraSpeed
MiMo-V2.6-Pro-UltraSpeed is the fast speed edition of Xiaomi's flagship found...
MiMo-V2-Flash
MiMo-V2-Flash is an open-source foundation language model developed by Xiaomi...
MiMo-V2-Omni
MiMo-V2-Omni is a frontier omni-modal model that natively processes image, vi...
MiMo-V2-Pro
MiMo-V2-Pro is Xiaomi's flagship foundation model, featuring over 1T total pa...
MiMo-V2.5
MiMo-V2.5 is a native omnimodal model by Xiaomi. It delivers Pro-level agenti...
MiMo-V2.5-Pro
MiMo-V2.5-Pro is Xiaomi’s flagship model, delivering strong performance in ge...
Cheapest hardware per quantisation.
Each row is one quantisation tier (the same weights compressed differently). Lower precision → lower VRAM → cheaper hardware, at the cost of small accuracy loss. $/Std. ist der günstigste aktuelle Preis. Ein Setup mit mehreren Karten nutzt eine echte Instanz dieser Größe, wenn eine vermietet wird; sonst ist es der Einzelkartenpreis mal Anzahl.
| Quantisation | Cheapest GPU config | Total VRAM | Live $/hr | tokens/sec | |
|---|---|---|---|---|---|
|
FP8
FP8 — 8-bit float (Hopper / Blackwell)
|
1536 GB |
$22.60/hr
Cyfuture AI · On-Demand
|
— | Compare → | |
|
INT4
INT4 — 4-bit integer (~4× VRAM saving)
|
640 GB |
$5.73/hr
geschätzt 8 × Einzelkartenpreis
Nosana · On-Demand
|
— | Compare → |
What it costs per month across providers.
Estimate your monthly bill for MiMo-V2.6-Pro across every host that publishes per-token pricing. Slide your token volumes; the chart + table re-rank cheapest-first.
Günstigster Anbieter links.
Gesamte Monatskosten — Eingabe- und Ausgabe-Tokens zusammen.
Rechnung im Detail.
| Anbieter | Monatssumme | |
|---|---|---|
| $6.04 | Registrieren ↗ | |
| $6.09 | Registrieren ↗ |
GPU mieten statt pro Token zahlen.
Bei einem Open-Weights-Modell wie MiMo-V2.6-Pro kannst du eine GPU mieten und die Inferenz selbst betreiben. Die Rechnung: günstigste GPU-Miete × 730 Stunden/Monat. Strom ist im Stundenpreis der Miete enthalten.
Angenommen wird, dass die GPU rund um die Uhr mit ca. 85 % Auslastung läuft. Bei schwankendem Traffic zahlst du für die API weniger und für die GPU wahrscheinlich mehr (Leerlaufstunden kosten trotzdem Miete). Die Break-even-Analyse findest du im Break-even-Tool Selbsthosting vs. API.
About MiMo-V2.6-Pro.
MiMo-V2.6-Pro is the lead checkpoint of Xiaomi's MiMo-V2.6 series, published as MiMo-V2.6-Pro-RL under the MIT licence. It is a sparse mixture-of-experts model with 1.02T total and 42B active parameters. It takes text, images, video and audio in one model, with a 1M-token context meant for long repositories, tool traces and multi-session agent runs. Xiaomi trained coding, general agent, visual and cybersecurity tasks in one mixed reinforcement learning run.