
MiniMax M2.5 Highspeed / 对话
MiniMax-M2.5-highspeed is an ultra-fast inference version of MiniMax M2.5. It maintains the full intelligent capability of the standard M2.5 model, featuring ultra-low latency, high throughput and outstanding cost performance. Built on advanced MoE architecture, it delivers rapid response speed while ensuring stable output quality. Optimized for high-concurrency business scenarios such as real-time dialogue, content generation and API service calls, it perfectly meets enterprise-level demands fo
Press Enter to add, Backspace to remove
Playground 对话
与 AI 模型开启对话,您可以询问任何问题。
AI 生成结果的准确性可能有所不同。
MiniMax M2.5 High SpeedInteractive Chat at High Throughput
MiniMax M2.5 High Speed is a latency-optimized chat sibling of M2.5 with streaming defaults, temperature and top-p sampling, stop sequences, and max output tokens up to 2048 — built for copilots that paint tokens as they arrive.

At a Glance
M2.5 High Speed Capabilities
Same chat completions contract as M2.5, tuned for interactive latency and progressive paint.
High-Speed Serving Path
A dedicated low-latency path for chat UIs, copilots, and turn-taking agents that cannot wait on batch queues. First tokens arrive early so the interface never feels stalled.

Streaming by Default
stream defaults to true so tokens arrive as they are generated. Consume SSE events for progressive chat paint without extra client plumbing.
Sampling Controls
Temperature default 0.7 (range 0–1) and top-p default 0.95 (range 0–1) give precise tone and diversity control without a sprawling parameter surface.
Bounded Completions
Stop sequences truncate trailing chatter and max_tokens caps length at 2048 for cost-aware product surfaces and parser-safe replies.
How It Works
Ship streaming chat in four focused steps.
POST Messages
Send system and user turns in the standard chat completions schema.
Stream Tokens
Consume SSE events immediately so the UI paints as the model thinks out loud.
Bound Output
Use stop sequences and max_tokens (1–2048) to keep replies parser-safe.
Tune Sampling
Adjust temperature and top-p to match brand voice without re-prompting.
High Speed Product Domains
Where every millisecond of first token matters.
Live Copilots
Inline suggestions with streaming first tokens.
Support Bots
Turn-taking chat with stop-bounded replies.
IDE Assist
Low-latency completions inside coding surfaces.
Ops Copilot
Fast internal Q&A over runbooks and tickets.
Prompt Tips
Keep high-speed replies tight and reliable.
High-speed paths shine on short, focused turns rather than multi-page dumps.
Put role, constraints, and output format in the system message.
Bound generation with stop when you will parse JSON or field lists.
MiniMax M2.5 High Speed Quickstart
Low-latency streaming chat completions.
curl -X POST "https://api.powertokens.ai/v1/chat/completions" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "MiniMax-M2.5-highspeed",
"messages": [
{"role": "user", "content": "Suggest three concise subject lines for a product launch email."}
],
"stream": true,
"temperature": 0.7
}'Technical Specifications
Confirmed parameters and runtime execution protocols.
价格详情
此模型的实际计费根据您在 API 请求中传入的具体参数动态计算。以下是具体的组合及其对应的价格:
| 模态 | 输入积分 | 输出积分 | 输入价格 | 输出价格 | 隐式缓存命中 | 显式缓存命中 | 创建缓存 |
|---|---|---|---|---|---|---|---|
| Standard | 570/ 1M Tokens | 2,280/ 1M Tokens | $0.570 | $2.280 | $0.029 29/ 1M Tokens | $0.029 29/ 1M Tokens | $0.357 357/ 1M Tokens |
Models from the Same Channel
Explore complementary models and alternative versions from the same provider channel.


MiniMax-M3.1-Flash
MiniMax-M3.1-Flash
Next-gen efficiency-focused multimodal model (428B MoE) for agentic workflows, coding, and computer use . Features a 1M-token context window and tunable thinking depth (low–max) . Scores 73.7% on DeepSWE v1.1, 61.7 on OSWorld V2, and 82% on CyberGym, while delivering 200+ TPS output speed at 70% lower agent-task cost than its predecessor


MiniMax M3
MiniMax-M3
Flagship multimodal foundation model supporting text, image, and video inputs with text output. Features a 1M-token context window via MiniMax Sparse Attention (MSA), cutting per-token compute to ~1/20 of previous gen at full context. Excels at long-horizon agentic work, coding, and tool use. Native multimodal training from step zero ensures deep semantic alignment. Scores 59.0% on SWE-Bench Pro and 83.5 on BrowseComp, surpassing Opus 4.7


MiniMax M2.7
MiniMax-M2.7
MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement. Built to actively participate in its own evolution, M2.7 integrates advanced agentic capabilities through multi-agent collaboration, enabling it to plan, execute, and refine complex tasks across dynamic environments.


MiniMax M2.7 Highspeed
MiniMax-M2.7-highspeed
A high-speed variant of MiniMax’s flagship M2.7 LLM, delivering 100 TPS—~60% faster than standard M2.7—with identical top-tier quality. It features a 204,800-token context window, near-Opus-level SWE performance, and recursive self-improvement. Optimized for low-latency coding, office tasks, and agent workflows, it balances extreme speed, reliability, and cost-efficiency for high-throughput real-world applications。
Recommended Related Models
Explore complementary video and multimodal models with your unified API key.


GLM-5.3
glm-5.3
Zhipu AI's flagship text model optimized for complex software engineering and long-horizon Agent tasks. Features a 1M-token context window with mandatory reasoning (3 levels: low/high/max). Coding capability improved 50% over GLM-5.2 on Z.ai Code Bench; scores SOTA on Terminal Bench 3.0. Excels in cybersecurity tasks (cyber vulnerability discovery)


MiMo-V2.6-Pro-UltraSpeed
mimo-v2.6-pro-ultraspeed
High-speed edition of Xiaomi's 1T-parameter flagship, delivering roughly 10x output speed while matching the original model's quality. Features a 1M-token context window and native multimodal input (text, image, video, audio). Optimized for latency-sensitive agentic workflows, coding, and long-horizon tasks


GLM-5.3 Flash
glm-5.3-flash
Z.ai's native multimodal model (320B-A18B) with text, image, video, and file input. Features hybrid sparse + linear attention, reducing attention compute 3x and KV cache 4.4x vs. GLM-5.3. Scores 57 on Artificial Analysis Intelligence Index—matching Claude Opus 4.8—with 1M context and 128K max output. Excels in agentic coding, vision-based Computer Use, and CAD reproduction


Qwen3.8 Flash
qwen3.8-flash
Alibaba's cost-efficient multimodal reasoning model. Supports text, image, and video inputs with text output. Features a native 1M-token context window for long documents, codebases, and agentic workflows. Excels in coding assistance, desktop interaction, chart analysis, and long-video understanding. Compatible with OpenAI/Anthropic protocols for seamless integration
Frequently Asked Questions
Everything you need to know before integrating this model.
Start Building with MiniMax M2.5 High Speed Today
Create an account in seconds to receive free test credits. No upfront payment or long-term contract required.