Skip to content
TrustList
MV
AI Model

MiMo-V2.6-Flash

New· 4

Xiaomi's smaller open-weight model for high-volume calls, accepting text, image, video and audio input

About MiMo-V2.6-Flash

The smaller of the two headline models in Xiaomi's MiMo-V2.6 series, MiMo-V2.6-Flash is built for work that is repeated many times rather than work that is hard once. Xiaomi announced it on 22 September 2026 and released the weights under the MIT licence.

309 billion parameters, 15 billion working at a time

Flash is a sparse mixture-of-experts model: 309 billion parameters overall, 15 billion of them active per token. Like the Pro model in the same series it takes a context of one million tokens, can write answers of up to 128,000 tokens, and accepts text, images, video and audio while returning text. Xiaomi aims it at workloads made of many frequent calls at scale, where a lower cost per request matters more than the last increment of capability.

Moving off mimo-v2.5

Xiaomi's overseas price list puts Flash in the same row as mimo-v2.5. That older model is due to be retired on 21 October 2026 at 10:00 Beijing time, which leaves a team calling mimo-v2.5 by name a limited window to move. The list also includes a cached-input rate and a batch interface billed at half the normal rates.

Weights and the API

Open weights went up on the model host on 21 September under the name MiMo-V2.6-Flash-RL. The card mentions that the model can also be reached through the API but, as with the Pro model, leaves open whether the two are identical, which is worth checking before relying on one to reproduce the other.

What is known about performance

Artificial Analysis had not published a page for the model when it was read on 22 September, so no independent score exists yet. Xiaomi reports its own results on DeepSWE, Terminal-Bench and OSWorld-Verified; these come from Xiaomi's own evaluation and are not verified figures.

A research companion

Released alongside it was MiMo-V2.6-Distill-Qwen-9B, a 9B model distilled onto an open base from another vendor. It comes as weights only, with no API, and Xiaomi offers it as a starting point for research rather than as a product.

TrustList summary

MiMo-V2.6-Flash is a language model from Xiaomi, released on 22 September 2026 as the smaller member of the MiMo-V2.6 series. Xiaomi positions it for high-frequency calls and large-scale work. It uses a mixture-of-experts design with 309 billion parameters in total and 15 billion active at a time, and its weights are published under the MIT licence.

What MiMo-V2.6-Flash does

  • Multimodal input: accepts text, image, video and audio, and returns text.
  • Long context: a window of one million tokens, with up to 128,000 output tokens per response.
  • Open weights: model weights released under the MIT licence.
  • Hosted access: the same model is also offered through an API.

Who it serves

Teams making high-frequency model calls or running large-scale work, as Xiaomi positions it.

Key facts

  • Developer: Xiaomi, as part of its MiMo-V2.6 series
  • Released: 22 September 2026
  • Licence: MIT
  • Context window: one million tokens

Summary written by TrustList from the business's own published information, Nov 2026.

Benchmarks & AI stats

BenchmarkOfficialCommunity avg
Artificial Analysis Intelligence Index v5.0.1llm27 index-

“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).

Sign in to add a benchmark score for this model.

Trust Score

4/ 100
Trust Score: New

An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.

Verification
0/100 · 20%
Reviews
0/100 · 30%
Awards
0/100 · 15%
Transparency
18/100 · 20%
Engagement
0/100 · 15%
Joining soon
Recommendations
Coming soon
Complaints
Coming soon

Updated 9/22/2026

What the business tells TrustList

Founded
2026 (0 years) evidence sent

Request a demo or quote from MiMo-V2.6-Flash

Protected by reCAPTCHA - Google Privacy Policy and Terms apply.

By sending, you agree we may share your request and contact details with the provider once you confirm your email.

Reviews

Write the first review of MiMo-V2.6-Flash

Used it? Your experience helps other buyers decide.

Write a review

Questions & answers

No questions yet. Be the first to ask about MiMo-V2.6-Flash.

MiMo-V2.6-Flash alternatives

Similar listings buyers compare against this one.

Aleph Alpha's open-weight English-German mixture-of-experts model for self-hosting, released under Apache 2.0 in October 2026

Microsoft AI's real-time speech-to-text model for voice agents, transcribing 60 languages with continuous language detection

Black Forest Labs' image generation and editing model with bounding-box layouts, multiple reference images and 4K output

Cloudflare's multimodal decision model, released 9 October 2026 under Apache 2.0: takes text, images, audio and video and returns a score for each allowed answer instead of free text. 30B mixture of experts with 3B active.

Cantina Security's first open-weight security model, a post-trained GLM-5.3-Flash for vulnerability research and exploit checks

Perplexity's open-weight decision model that picks among given options for routing and checks in AI systems

Smaller, latency-optimised Cloudflare decision model that returns probability-scored classifications of text or images

Microsoft AI's text-to-speech model for keeping one voice across 23 languages, with voice cloning from short audio