MiniMax M2 is a state-of-the-art agent-based reasoner designed for high efficiency and advanced autonomous capabilities. It leverages a Mixture of Experts (MoE) architecture with 230 billion total parameters, activating 10 billion parameters per inference for optimal performance. This model excels in latency reduction and throughput, making it suitable for demanding AI applications.
vs Claude Opus: MiniMax M2 emphasizes swift, scalable agent reasoning with coding-centered capabilities, whereas Claude Opus targets maximal reasoning depth and complex multi-step problem solving with robust tool and memory integration.
vs Gemini 2.5: MiniMax M2 surpasses Gemini 2.5 Pro in overall intelligence scores, placing it among the top five models globally, and exceeds it in efficiency due to its MoE architecture activating only a small subset of parameters. Gemini 2.5 excels in multimodal creativity, generating images, art, and audio, whereas MiniMax M2 is more focused on reasoning, coding, and autonomous agent tasks.
vs PaLM 2: PaLM 2 is renowned for its multilingual, coding, and research capabilities, but MiniMax M2 outperforms it in open intelligence rankings and autonomous reasoning benchmarks. PaLM 2’s versatility makes it suitable for a wide range of applications, whereas MiniMax M2’s strength lies in cost-efficient, high-performance agent reasoning for specialized tasks.
vs GPT-4: GPT-4 remains the industry leader for multilingual understanding and a wide array of domains beyond coding and reasoning, while MiniMax M2 specifically optimizes those domains tied to autonomous agent reasoning, coding, and deep search functions.
Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.
import requests r = requests.post( "https://api.aimlapi.com/v1/chat/completions", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "minimax/m2", "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", "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","messages":[{"role":"user","content":"Hello!"}]}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| Humanity's Last Exam | 12.5% | Expert-level questions across many domains | Source | July 8, 2026 |
| LiveCodeBench | 83% | Contamination-free competitive programming problems | Source | July 8, 2026 |
| AIME 2025 | 78% | Competition mathematics (AIME), 2025 | Source | July 8, 2026 |
| GPQA Diamond | 78% | Google-proof graduate science questions (hardest subset) | Source | July 8, 2026 |
| Terminal-Bench | 46.3% | Autonomous shell/terminal task completion | Source | July 8, 2026 |
| SWE-bench Verified | 69.4% | Resolving verified real GitHub issues | Source | July 8, 2026 |
| Intelligence | 18.6 | Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Math | 78.3 | Composite score across standardised mathematics evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
MiniMax M2 This page | Coding + agents | |||
| Reasoning + agents | ||||
| Balanced coding + agents | ||||
| Long-context, multimodal & agentic workflows | ||||
| Reasoning + agents |
MiniMax M2 has a 1,000,000 tokens context window and can return up to 40,960 tokens.
MiniMax M2 takes text as input and returns text.
Use minimax/m2 as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
MiniMax M2 is priced at input $0.39 / 1M tokens, output $1.56 / 1M tokens.
Yes, MiniMax M2 can stream responses as they are generated.
MiniMax M2 was built by MiniMax.
