Submit
XiaomiOpen-weight·1000K context·— params·MIT

MiMo-V2.6-Proxiaomi/mimo-v2-6-pro

Xiaomi's September 22 multimodal reasoning model, available through the MiMo API and downloadable MIT-licensed MOPD weights. The standard API and the separately priced UltraSpeed route require different budgets. Evaluate tool-call behavior on your own tasks before migrating.

Cheapest blended:$0.54 / 1M tokenson Xiaomi MiMo · 1 provider listed

Official sources checked October 2, 2026. Evaluation suggestions below are not benchmark results.

Standard, UltraSpeed and batch are different budgets

The matrix shows the standard USD pay-as-you-go route. The separate mimo-v2.6-pro-ultraspeed ID costs $4.35 input, $0.036 cached input and $8.70 output per million tokens—ten times the standard rates. Its latency advantage is a provider claim; no independent timing test was performed here.

Batch requests have a separate half-price schedule: $0.2175 input, $0.0018 cached input and $0.435 output per million tokens. Batch does not apply to UltraSpeed. A queue that can tolerate asynchronous completion should compare batch total cost and deadline reliability before paying for faster interactive serving.

Migration and checkpoint selection

The linked MOPD weights use an MIT license. Xiaomi describes MOPD as addressing repeated tool calls; that does not prove the problem is absent in every agent harness. Pin the checkpoint or API ID, preserve your old configuration, and compare repeated calls, failed tasks and total output tokens on a fixed task set.

The older MiMo-V2.5 and V2.5-Pro APIs are scheduled to retire on October 21 at 10:00 Beijing time. This is a future retirement notice, not evidence they are already unavailable. Check the provider's current notice before setting a migration deadline.

A useful Pro versus Flash evaluation

Choose tasks where a wrong action is expensive: multi-step code changes, tool selection from similar schemas, and a decision that needs evidence from several input files. Record acceptable completion, unintended actions, attempts and billable usage. Use those results to decide whether Pro's higher token rate buys a useful improvement.

The advertised 1M context limit is not a demonstrated retrieval score. Put the necessary evidence near the start, middle and end of a long input and score omissions separately. Keep the input format, tool budget and acceptance criteria fixed across routes.

Official source 1 →Official source 2 →Official source 3 →Official source 4 →

Pricing across providers

Sort by:
ProviderInput /1MOutput /1MBlended /1MLatency p50FormatFreshnessAction
Xiaomi MiMo
mimo-v2.6-pro

Standard USD pay-as-you-go rate. UltraSpeed and batch prices differ; verify account and region eligibility.

$0.435
Cached $0.0036
$0.87$0.544—OpenAI-compatibleNeeds 24d agoTry →

Affiliate disclosure: We may earn a commission from qualified signups. Pricing independence is enforced at the data layer — see our Editorial Independence Policy.

Works with

Point any of these clients at a hosting's base URL — they all speak at least one of this model's endpoint protocols (OPENAI_COMPATIBLE).

Capabilities

  • reasoning
  • coding
  • tool_use
  • multimodal_input

Code samples

Example using Xiaomi MiMo — the cheapest hosting for this model as of last verification. Swap base_url and model to use a different provider from the matrix above.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.xiaomimimo.com/v1",
)

response = client.chat.completions.create(
    model="mimo-v2.6-pro",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

Technical specs

Context
1000K
Max output
128K
Parameters
—
Release
2026-09-22
Training cutoff
—
License
MIT
Hugging Face →Official documentation →

Similar models

Compare with

  • MiMo-V2.6-Pro vs DeepSeek V4.1 Flash
    Comparison planned — not yet published
  • MiMo-V2.6-Pro vs GLM 5.3 Flash
    Comparison planned — not yet published
  • MiMo-V2.6-Pro vs Kimi K2.5
    Comparison planned — not yet published

Frequently asked

How much does MiMo-V2.6-Pro cost?+
The cheapest public hosting is $0.54 per 1M blended tokens on Xiaomi MiMo. 1 total providers are listed above with per-input / per-output / cached pricing.
Is MiMo-V2.6-Pro open-source? Can I fine-tune it?+
MiMo-V2.6-Pro has published weights. License: MIT. Review the linked repository and license for deployment, fine-tuning and commercial-use conditions.
Is MiMo-V2.6-Pro OpenAI-compatible?+
At least one listed hosting exposes an OpenAI-compatible API, so you can point an existing openai SDK client at the Provider's base_url and use the Provider's model name. See the Code Samples above for a copy-pasteable example.
What's the maximum context window for MiMo-V2.6-Pro?+
The model supports up to 1,000,000 tokens of context (input + output). Some hosted versions may impose a smaller limit — check the "Context" column in the pricing matrix for each provider.