Name: MiniMax M2-her API
Brand: MiniMax

Overview

What Is MiniMax M2-her API?

MiniMax M2-her is a specialized conversational model optimized for role-based interaction and immersive dialogue systems. Unlike general-purpose LLMs, it is engineered to maintain coherence across extended exchanges, even when multiple characters, shifting contexts, or layered storylines are involved.

The model places strong emphasis on world consistency, narrative flow, and user preference adaptation, making it suitable for interactive fiction, AI companions, and structured storytelling systems. It is designed to reduce common long-context failures such as reference confusion, repetitive dialogue loops, and inconsistent character behavior.

API Pricing

Input: $0.39 / 1M tokens
Output: $1.56 / 1M tokens

System Behavior Overview

Capability Area	Description	Impact on Experience
Character consistency	Maintains stable personalities across long sessions	More believable roleplay and dialogue
Context memory handling	Tracks long conversation history effectively	Reduced repetition and confusion
Story progression	Actively advances narrative flow	More dynamic interactions
Multi-role support	Handles multiple characters in one session	Enables complex storytelling setups

Key Characteristics

Character Consistency Across Long Interactions

MiniMax-M2-her is distinguished by its ability to maintain character identity across long and complex interactions. In extended role-play scenarios, it reduces the typical “personality drift” seen in many large language models, helping fictional characters remain consistent and believable.

Multi-Character Dialogue Management

It also supports multi-character environments, where several distinct voices may exist within a single conversation. This includes situations where the model alternates between narrators, supporting characters, and direct user interaction while maintaining clarity and structure.

Long-Horizon Narrative Stability

Another defining feature is long-horizon stability. The model is engineered to preserve coherence across very long sessions, making it suitable for interactive storytelling experiences that extend far beyond typical chatbot exchanges.

Emotionally Aware Response Behavior

Emotionally aware dialogue handling is also central to its behavior. The model interprets subtle conversational cues such as hesitation, emotional intensity, or implied intent, and adjusts pacing, tone, and narrative direction accordingly.

Technical Overview

Although MiniMax-M2-her is not positioned as a reasoning or coding-focused model, its underlying architecture is built for scalable conversational simulation. It leverages large-scale dialogue datasets that include multi-turn interactions, character-driven scenarios, and structured narrative environments.

Training methodologies emphasize consistency over time, often using self-play style conversation generation where multiple AI agents interact within the same scenario. This allows the model to learn how narratives evolve naturally rather than in isolated exchanges.

Performance Profile

MiniMax-M2-her is strongest in areas that require immersion and sustained interaction quality. It excels in maintaining role-play integrity, preserving character consistency, and handling long conversational arcs without losing coherence.

Its narrative creativity is also a key strength, allowing it to generate dynamic and evolving storylines that feel responsive to user input. However, it is not optimized for analytical tasks such as complex reasoning, coding, or strict factual problem-solving, where other model families typically perform better.

Performance Characteristics

While MiniMax does not position M2-her as a reasoning-first or coding-first model, it excels in:

Capability Area	Performance Profile
Role-play immersion	★★★★★
Character consistency	★★★★★
Long conversation stability	★★★★★
Narrative creativity	★★★★☆
Coding / analytical reasoning	★★☆☆☆

The model is explicitly optimized for interaction quality over deterministic task solving, making it distinct from productivity-focused LLMs.

Strengths and Limitations

The primary strength of MiniMax-M2-her lies in its ability to sustain immersive, emotionally coherent interactions over long periods. It delivers strong character consistency, natural narrative pacing, and smooth multi-voice dialogue management.

Its limitations emerge in domains outside storytelling. It is not intended for high-precision reasoning tasks, structured enterprise automation, or technical problem-solving. In such contexts, its strengths in narrative generation may become less relevant compared to more analytically focused models.

‍

Uptime and performance

Uptime (7d): 100%

Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.

Use cases

Designed for role-playing and multi-turn conversations, with a dialogue-first focus on expressive back-and-forth interaction.

Suited to immersive roleplay and character-driven chat, where maintaining tone and personality across exchanges matters.

Used for storytelling and story-driven conversations, including scenarios that benefit from natural flow and vivid interaction.

How to use MiniMax M2-her API

Install any OpenAI-compatible SDK, point it at api.aimlapi.com/v1, and set the model to minimax/m2-her.

import requests

r = requests.post(
    "https://api.aimlapi.com/v1/chat/completions",
    headers={"Authorization": "Bearer " + AIMLAPI_KEY},
    json={
      "model": "minimax/m2-her",
      "messages": [
        {
          "role": "user",
          "content": "Hello!"
        }
      ]
    },
)
print(r.json())

const r = await fetch("https://api.aimlapi.com/v1/chat/completions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "minimax/m2-her",
    "messages": [
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
  }),
});
console.log(await r.json());

curl -X POST https://api.aimlapi.com/v1/chat/completions \
  -H "Authorization: Bearer $AIMLAPI_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"minimax/m2-her","messages":[{"role":"user","content":"Hello!"}]}'

OpenAI-compatible — swap the base URL and it works with your existing SDK.