Skip to main content

MiMo V2 Flash

**MiMo-V2-Flash** is a Mixture-of-Experts (MoE) language model with **309B total parameters** and **15B active parameters**. Designed for high-speed reasoning and agentic workflows, it utilizes a novel hybrid attention architecture and Multi-Token Prediction (MTP) to achieve state-of-the-art performance while significantly reducing inference costs.

Input / output price
$0.1078 / $0.28 per 1M
Context
262K tokens
Released
Dec 16, 2025
Agent setup

Paste into Claude Code, Cursor, or Codex

Help me set up the Respan gateway: fetch and follow https://www.respan.ai/docs/agent-setup.md. Use model "mimo-v2-flash".

Read docs

Providers

One provider serves this model, priced per 1M tokens after gateway discounts.

ProviderInputOutput
Featherlessfeatherless/mimo-v2-flash
$0.1078/M$0.28/M

Pricing

USD per 1M tokens on Featherless, after gateway discounts.

Input
$0.1078 per 1M
Output
$0.28 per 1M
Cache read
$1.00 per 1M

Cost by volume

Total cost from 0 to 1T tokens.

$0.1078/MFeatherless

Benchmarks

The Respan Index plus 8 published scores from 2 sources, each linked to where it was published.

23.0/ 100

Respan Index#113 of 176

176 ranked models. The shaded band is this model's 95% interval, 15.2 to 41.4. Gemini 2.5 Pro and Qwen3.5-Flash fall inside it, so the gap is within the margin of error.
MiMo V2 Flash by category: Coding 6.6, Agents 20.9 (estimated), Reasoning 27.7, Knowledge & language 19.9, Human preference 48.0
  1. 113MiMo-V2-Flash23.0
  2. 112Gemini 2.5 Pro23.2
  3. 114Qwen3.5-Flash22.7

Published scores

BenchmarkScore
SWE-bench Verifiedhuggingface.co73.4%
Terminal-Bench 2.0huggingface.co38.5%

Compare MiMo V2 Flash: MiMo-V2-Flash vs MiMo-V2.6-Flash, MiMo-V2-Flash vs MiMo-V2.6-Pro, MiMo-V2-Flash vs MiMo-V2.5-Pro

Apps

Use MiMo V2 Flash in the tools you already run. Each guide points the app at the Respan gateway.

Setup guides for more apps are in the integration docs.

MiMo V2 Flash FAQ

Short answers, built from the data on this page.

MiMo V2 Flash is a model from Xiaomi, released on Dec 16, 2025. **MiMo-V2-Flash** is a Mixture-of-Experts (MoE) language model with **309B total parameters** and **15B active parameters**. Designed for high-speed reasoning and agentic workflows, it utilizes a novel hybrid attention architecture and Multi-Token Prediction (MTP) to achieve state-of-the-art performance while significantly reducing inference costs.

On the Respan gateway, MiMo V2 Flash costs $0.1078 per 1M input tokens and $0.28 per 1M output tokens on Featherless, after gateway discounts.

MiMo V2 Flash is served by Featherless.

MiMo V2 Flash accepts up to 262K tokens of context.

Point any OpenAI-compatible SDK at https://gateway.respan.ai/v1 with your Respan API key and set the model to "mimo-v2-flash". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

MiMo V2 Flash scores 23.0 out of 100 on the Respan Index v1.1, ranking #113 of 176 models. By category: Coding 6.6, Agents 20.9 (estimated), Reasoning 27.7, Knowledge & language 19.9, and Human preference 48.0. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by Respan.

MiMo V2 Flash has 8 published benchmark scores, including SWE-bench Verified 73.4%, GPQA Diamond 83.7%, and AIME 2025 94.1%. Each score links to where it was published; the data is collected by Respan.

Certified where it counts.

Built to meet the security and privacy standards that enterprise and healthcare teams require.

ISO 27001

The internationally recognized standard for information security management.

SOC 2

Secure, compliant management of your data across all of our systems.

GDPR

Operated under GDPR, the world's strictest standard for data privacy.

HIPAA

HIPAA compliant, with a BAA available for healthcare teams.

Works with your entire stack

Use Respan with your favorite frameworks and tools.

AnthropicAI21 LabsxAIVercel AI SDK
LangChainOpenRouterAssemblyAICursor
LiteLLMGroqBasetenMem0
CohereCogneeLlamaIndexFireworks
DeepSeekLinkupHaystackTogether AI
PerplexityInference.netPostHogRubyLLM
SlackMastraAzure OpenAIMistral
AWS BedrockMoonshotHyperspellBAML
ModaInstructorGoogle GeminiParasail
ReductoZapierAgnoNebius
OpenAINovita AIReplicateSuperagent

Built for AI agents.
Break less.
Ship more.