MiniMax: MiniMax M2.7 API

minimax-m2.7

MiniMax M2.7 API is a text-only agent model with a 205K context and broad tool and structured-output controls. It is a smaller-context, lower-priced alternative to MiniMax M3. ^[1]

MiniMax M2.7 API has no exact mapped result in LLM2014 logic 2026-08. Historical scores and scores for other versions are not presented as current results for this model. ^[1][2]

MODALITIES

Price$0.18 / $0.72 / 1MCONTEXT204.8K

curl https://apineed.com/v1/chat/completions \
  -H "Authorization: Bearer $API_NEED_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"minimax-m2.7"}'

Playground

Try MiniMax M2.7 in the API Need Playground.

Provider

Current API Need routing and live API pricing. ^[1]

API Need25% off

Uptime

Total Context

204.8K

Max Output

131.072K

Input

$0.18/ 1M

Output

$0.72/ 1M

Cache Read

Route priced

Cache Write

Route priced

Pricing

Pricing uses the current live API Need configuration for this model.

Official

Provider baseline

Input $0.24/ 1MOutput $0.96/ 1M

Baseline

APINEED25% off

Prepaid credits

Input $0.18/ 1MOutput $0.72/ 1M

Availability

Recent public performance data for this model.

API statusLive metrics

API Need gateway99.50% uptime

Benchmarks

Editorial benchmark data is shown when this model has a verified match. ^[2]Intelligence uses the source’s median-score ranking. Efficiency views are APINEED calculations, not official LLM2014 rankings. Test costs and reference prices are converted at ¥7 per US dollar; they are not APINEED selling prices. Missing values are N/A. These results describe the source’s tested configurations, not guaranteed APINEED endpoint performance.


1st	GPT-5.5 (xhigh)	80.23	73.89	7.9%	534s	33,219	$27.51	$29.57
2nd	GPT-5.6 Sol (xhigh)	78.37	69.42	11.42%	381s	17,769	$14.71	$29.57
3rd	Kimi-K3 (max)	75.77	67.66	10.7%	1144s	41,432	$17.40	$15.00
4th	Claude Opus 5 (xhigh)	71.24	64.74	9.12%	437s	26,705	$18.43	$24.64
5th	GLM-5.3 (max)	74.48	63.91	14.19%	1009s	57,408	$6.43	$4.00
6th	GLM-5.3-Flash (max)	69.55	60.52	12.98%	863s	38,787	$0.22	$0.20
7th	Qwen3.8-Max (xhigh)	64.68	60.05	7.16%	1297s	65,135	$9.38	$5.14
8th	DeepSeek V4 Pro 0813(max)	70.13	59.63	14.97%	1601s	73,350	$7.92	$3.86
9th	DeepSeek-V4-Flash-Vision-Exp (max)	62.86	58.10	7.57%	974s	79,420	$2.86	$1.29
10th	Gemini 3.7 Flash (high)	66.62	57.56	13.6%	131s	26,596	$2.75	$3.70
11th	Grok 4.6 (high)	67.93	56.41	16.96%	813s	33,807	$5.60	$5.91
12th	Gemini 3.1 Pro (high)	69.35	55.84	19.48%	235s	28,338	$9.39	$11.83
13th	DeepSeek V4 Flash 0731 (max)	66.34	55.23	16.75%	881s	74,657	$2.69	$1.29
14th	GLM-5.2 (max)	66.53	54.70	17.78%	891s	46,273	$5.18	$4.00
15th	Qwen3.8-Flash (xhigh)	65.44	54.49	16.73%	844s	64,942	$0.70	$0.39
16th	GPT-5.6 Luna (xhigh)	65.92	51.62	21.69%	247s	37,885	$1.25	$1.18
17th	Qwen3.8-27B (xhigh)	58.48	47.65	18.52%	2318s	73,987	$3.55	$1.71
18th	Doubao-Seed-2.1-pro (high)	55.12	45.78	16.94%	2026s	85,238	$10.23	$4.29
19th	Muse Spark 1.2 (xhigh)	57.93	45.14	22.08%	359s	47,209	$5.54	$4.19
20th	Gemini 3.7 Flash (low)	53.50	44.06	17.64%	59s	10,673	$1.10	$3.70
21st	Claude Sonnet 5 (xhigh)	47.80	39.57	17.22%	545s	45,066	$12.44	$9.86
22nd	Tencent Hy3 (high)	57.89	39.51	31.75%	1076s	53,214	$0.85	$0.57
23rd	Qwen3.7-Plus (high)	51.36	38.74	24.57%	1254s	58,610	$1.88	$1.14
24th	MiniMax-M3	43.83	37.37	14.74%	982s	57,486	$1.93	$1.20
25th	Doubao-Seed-2.0-lite 0428 (high)	34.13	26.18	23.29%	669s	27,845	$0.40	$0.51
26th	Claude Opus 5	34.94	24.58	29.65%	156s	10,261	$7.08	$24.64
27th	Gemini 3.5 Flash Lite (high)	29.51	20.44	30.74%	192s	17,842	$1.23	$2.46
28th	Ling-3.0-flash	28.37	17.05	39.9%	735s	95,538	N/A	N/A
29th	Gemma 4 31B	21.36	16.09	24.67%	756s	15,290	$0.17	$0.39
30th	openPangu-2.0-Pro	20.02	15.75	21.33%	645s	28,215	$1.64	$2.07
31st	Qwen3.7-Plus	20.84	15.63	25%	323s	8,910	$0.29	$1.14
32nd	GPT-5.5 Instant	21.73	14.09	35.16%	28s	1,712	$1.42	$29.57
33rd	Mistral Medium 3.5	17.25	13.87	19.59%	200s	28,129	$5.82	$7.39
34th	Step-3.7-Flash	26.18	13.74	47.52%	282s	47,180	$1.53	$1.16
35th	MiMo-V2.5-Pro	26.31	13.00	50.59%	550s	33,095	$0.79	$0.86
36th	Dots3-Note Preview	18.63	12.95	30.49%	210s	28,386	N/A	N/A
37th	ERNIE 5.1	16.32	11.32	30.64%	468s	24,522	$1.77	$2.57
38th	Qwen3.8-27B	16.48	10.61	35.62%	192s	8,914	$0.43	$1.71
39th	openPangu-2.0-Flash	19.24	10.16	47.19%	583s	29,052	$0.19	$0.23
40th	LongCat-2.0	17.61	9.74	44.69%	374s	14,273	$0.46	$1.14
41st	Claude Sonnet 5	16.49	9.67	41.36%	113s	6,175	$1.70	$9.86
42nd	GLM-5.2	12.42	8.05	35.19%	24s	929	$0.10	$4.00
43rd	DeepSeek V4 Flash 0731	13.57	7.93	41.56%	53s	4,621	$0.17	$1.29
44th	Gemini 3.1 Flash Lite	9.01	7.11	21.09%	11s	707	$0.03	$1.48
45th	Doubao-Seed-2.1-pro	13.03	6.17	52.65%	319s	6,948	$0.83	$4.29
46th	Mistral Medium 3.5	8.45	5.93	29.82%	18s	2,589	$0.54	$7.39

Quick Start

Call MiniMax M2.7 with its real backend model ID.

Get an API key

Create an API key in the product app.

export API_NEED_API_KEY=sk-apineed-v1-...

Send a request

Call /v1/chat/completions with model minimax-m2.7.

TypeScript SDKPythoncURLOpenAI SDK

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.API_NEED_API_KEY,
  baseURL: "https://apineed.com/v1"
});

const result = await client.chat.completions.create({
  model: "minimax-m2.7",
  messages: [
    { role: "user", content: "Why is the sky blue?" }
  ]
});

console.log(result.choices[0].message.content);

Handle the response

Validate responses and errors in your application.

curl https://apineed.com/v1/chat/completions \
  -H "Authorization: Bearer $API_NEED_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "minimax-m2.7",
  "stream": true,
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ]
}'

Endpoints

Endpoints currently returned by the live pricing API.

POST/v1/chat/completions

Authorization: Bearer $API_NEED_API_KEY
Content-Type: application/json
HTTP-Referer: optional - your site URL, for rankings
X-Title: optional - your site name, for rankings
Model: minimax-m2.7

Parameters

Supported parameters are shown when verified metadata is available. ^[1]

No verified parameter metadata is currently available.

MiniMax M2.7 API Q&A

Model-specific answers for teams comparing MiniMax M2.7 API on capability, benchmark evidence, integration, and cost.

Does MiniMax M2.7 API have a current benchmark rank?

MiniMax M2.7 has no exact mapped result in LLM2014 logic 2026-08. Historical scores and scores for other versions are not presented as current results for this model.

How large is the MiniMax M2.7 API context window?

The catalog lists 204,800 tokens of context with text input and text output.

What does MiniMax M2.7 API cost through APINEED?

Input is $0.18 and output is $0.72 per million tokens after the 25% discount.

Which MiniMax M2.7 workflows should be tested?

For MiniMax M2.7 API, test coding, debugging, tool execution, and document generation with realistic context and explicit completion limits.

Which endpoint serves MiniMax M2.7 API?

Call MiniMax M2.7 API through https://apineed.com/v1/chat/completions with minimax-m2.7 as the model value. Existing OpenAI SDK clients usually need only the APINEED base URL and key.

How do I validate MiniMax M2.7 API before release?

Evaluate MiniMax M2.7 API on representative prompts, record quality, latency, and token use, then choose reasoning settings and fallbacks from those results.

How to deploy MiniMax M2.7

Connect the live model to your application.

Top up

Add usage credit in the product app.

Create a key

Create and securely store an API key.

Configure the endpoint

Send requests to /v1/chat/completions.

Monitor responses

Track status, latency, and errors in your application.

More live models

MiniMax M3minimax/minimax-m3 H3minimax/minimax-h3