{"id":21129,"date":"2026-10-27T00:24:25","date_gmt":"2026-10-27T04:24:25","guid":{"rendered":"https:\/\/wp.glbgpt.com\/?p=21129"},"modified":"2026-10-27T00:45:45","modified_gmt":"2026-10-27T04:45:45","slug":"minimax-m3-1-flash-guide","status":"publish","type":"post","link":"https:\/\/wp.glbgpt.com\/jp\/hub\/minimax-m3-1-flash-guide","title":{"rendered":"MiniMax M3.1 \u30d5\u30e9\u30c3\u30b7\u30e5\u30ac\u30a4\u30c9\uff1aAPI\u3001\u4fa1\u683c\u3001\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3001\u30d1\u30e9\u30e1\u30fc\u30bf\u3001\u304a\u3088\u3073\u7121\u6599\u30a2\u30af\u30bb\u30b9"},"content":{"rendered":"<section style=\"box-sizing:border-box;margin:24px 0;padding:28px;border:1px solid #d8e6df;border-radius:24px;background:linear-gradient(135deg,#effaf4 0%,#fff5f2 100%);font-family:system-ui,-apple-system,Segoe UI,sans-serif;color:#20332d\">\n  <div style=\"display:inline-block;padding:7px 12px;border-radius:999px;background:#d8eee1;color:#2b6a4b;font-size:12px;font-weight:800;letter-spacing:.08em;text-transform:uppercase\">\u7c21\u5358\u306a\u56de\u7b54<\/div>\n  <h2 style=\"margin:15px 0 10px;font-size:clamp(25px,4vw,38px);line-height:1.12\">A fast, low-cost model for long-context agent work<\/h2>\n  <p style=\"margin:0;max-width:850px;font-size:17px;line-height:1.7\">MiniMax M3.1 Flash launched on October 19, 2026 with a 1M-token context window, adjustable thinking effort, multimodal input, and tool support. We tested three real tasks\u2014exact JSON, JavaScript debugging, and HTTP 503 diagnosis\u2014and all three completed successfully.<\/p>\n  <div style=\"display:grid;grid-template-columns:repeat(auto-fit,minmax(170px,1fr));gap:12px;margin-top:20px\">\n    <div style=\"padding:17px;border-radius:16px;background:#fff;box-shadow:0 8px 20px #416e5518\"><div style=\"font-size:13px;color:#6e8178\">\u6587\u8108<\/div><div style=\"margin-top:5px;font-size:29px;font-weight:850;color:#2b6a4b\">1M<\/div><div style=\"font-size:13px;color:#6e8178\">\u30c8\u30fc\u30af\u30f3<\/div><\/div>\n    <div style=\"padding:17px;border-radius:16px;background:#2f4b42;color:#fff;box-shadow:0 8px 20px #416e5526\"><div style=\"font-size:13px;color:#d3ebdc\">\u6295\u5165\u4fa1\u683c<\/div><div style=\"margin-top:5px;font-size:29px;font-weight:850\">$0.15<\/div><div style=\"font-size:13px;color:#d3ebdc\">100\u4e07\u30c8\u30fc\u30af\u30f3\u3042\u305f\u308a<\/div><\/div>\n    <div style=\"padding:17px;border-radius:16px;background:#fff;box-shadow:0 8px 20px #416e5518\"><div style=\"font-size:13px;color:#6e8178\">Real tasks<\/div><div style=\"margin-top:5px;font-size:29px;font-weight:850;color:#aa6570\">3\/3<\/div><div style=\"font-size:13px;color:#6e8178\">\u5b8c\u4e86<\/div><\/div>\n  <\/div>\n<\/section>\n\n\n\n<p class=\"wp-block-paragraph\">If you are comparing low-cost agent models for coding, tool use, or long documents, this guide puts the useful facts in one place: what the model can do, what the official scores actually mean, how to call it, how to estimate cost, and where access is currently visible. You can also compare the broader <a href=\"https:\/\/www.glbgpt.com\/hub\/best-ai-model-for-coding\/\">\u30b3\u30fc\u30c7\u30a3\u30f3\u30b0\u306b\u6700\u9069\u306aAI\u30e2\u30c7\u30eb<\/a> list before choosing a default.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/www.glbgpt.com\/home?inviter=hub_popup&amp;login=1\"><strong>Try MiniMax M3.1 Flash on GlobalGPT<\/strong><\/a> through the general model workspace. The authenticated test route used for this article exposed the model as \u201cMiniMax M3.1 Flash\u201d; platform entitlements and model availability can change by account and region.<\/p>\n\n\n\n<div class=\"wp-block-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<figure class=\"wp-block-image size-large\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1024\" height=\"645\" src=\"https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-1024x645.png\" alt=\"globalgpt \u30c0\u30c3\u30b7\u30e5\u30dc\u30fc\u30c9\" class=\"wp-image-19280\" srcset=\"https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-1024x645.png 1024w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-300x189.png 300w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-18x11.png 18w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-767x483.png 767w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-1536x967.png 1536w, https:\/\/wp.glbgpt.com\/wp-content\/uploads\/2026\/07\/glb-2048x1289.png 2048w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<div class=\"wp-block-buttons is-content-justification-center is-layout-flex wp-container-core-buttons-is-layout-3e41869c wp-block-buttons-is-layout-flex\">\n<div class=\"wp-block-button\"><a class=\"wp-block-button__link has-black-color has-text-color has-background has-link-color wp-element-button\" href=\"https:\/\/www.glbgpt.com\/home?inviter=hub_popup&amp;login=1\" style=\"background:linear-gradient(135deg,rgba(250,183,0,0.55) 0%,rgba(255,106,0,0.66) 96%)\"><strong>GlobalGPT\u3067100\u7a2e\u985e\u4ee5\u4e0a\u306e\u30c8\u30c3\u30d7\u30e2\u30c7\u30eb\u3092\u304a\u8a66\u3057\u304f\u3060\u3055\u3044<\/strong><\/a><\/div>\n<\/div>\n<\/div>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a href=\"#what-is-minimax-m31-flash\">What is MiniMax M3.1 Flash?<\/a><\/li>\n\n\n\n<li><a href=\"#benchmarks\">Official benchmarks and speed claims<\/a><\/li>\n\n\n\n<li><a href=\"#pricing\">API pricing and cost examples<\/a><\/li>\n\n\n\n<li><a href=\"#api-setup\">API setup: OpenAI-compatible and Anthropic-compatible<\/a><\/li>\n\n\n\n<li><a href=\"#parameters-and-modalities\">Parameters, thinking, tools, and multimodal limits<\/a><\/li>\n\n\n\n<li><a href=\"#hands-on-tests\">What our GlobalGPT tests returned<\/a><\/li>\n\n\n\n<li><a href=\"#free-access\">Free access, M Plan, and MiniMax Code<\/a><\/li>\n\n\n\n<li><a href=\"#m31-vs-m3\">M3.1 Flash vs M3<\/a><\/li>\n\n\n\n<li><a href=\"#who-should-use-it\">\u3069\u306e\u3088\u3046\u306a\u4eba\u306b\u9069\u3057\u3066\u3044\u307e\u3059\u304b\uff1f<\/a><\/li>\n\n\n\n<li><a href=\"#verdict\">\u8a55\u6c7a<\/a><\/li>\n\n\n\n<li><a href=\"#faq\">\u3088\u304f\u3042\u308b\u3054\u8cea\u554f<\/a><\/li>\n<\/ul>\n\n\n\n<h2 id=\"what-is-minimax-m31-flash\" class=\"wp-block-heading\">What is MiniMax M3.1 Flash?<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">MiniMax M3.1 Flash is the company\u2019s fast agent-oriented model for coding, computer-use workflows, tool calling, cybersecurity tasks, and long-context reasoning. The <a href=\"https:\/\/www.minimax.io\/blog\/minimax-m31f\">\u767a\u58f2\u306e\u304a\u77e5\u3089\u305b<\/a> dates the release to October 19, 2026, while the <a href=\"https:\/\/platform.minimax.io\/docs\/guides\/text-generation.md\">official invocation guide<\/a> lists a 1M-token context window.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The word \u201cFlash\u201d describes the model\u2019s speed and cost position, not a separate lightweight chat-only mode. Thinking is enabled by default and cannot be switched off in the documented API behavior. You can select a thinking effort level, which gives you a practical trade-off between response depth, latency, and output cost.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">That makes the model interesting for repository tasks, structured planning, API troubleshooting, and agents that need to keep a large amount of context in one turn. It is still worth matching the model to the workload: for a useful model-selection framework, see this <a href=\"https:\/\/www.glbgpt.com\/hub\/best-ai-models\/\">AI\u30e2\u30c7\u30eb\u306e\u9078\u3073\u65b9\u30ac\u30a4\u30c9<\/a> and compare the specific tool or coding constraints you have.<\/p>\n\n\n\n<h2 id=\"benchmarks\" class=\"wp-block-heading\">Official benchmarks and speed claims<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The launch article publishes the following scores and positioning. They are <strong>MiniMax-reported<\/strong> figures, not independent reruns from this article. The right way to read them is as the provider\u2019s release evidence and task coverage, not as a promise that every prompt will match a leaderboard result.<\/p>\n\n\n\n<section style=\"box-sizing:border-box;margin:26px 0;padding:24px;border:1px solid #d7e3dc;border-radius:20px;background:linear-gradient(135deg,#f2faf5,#fff8f4);font-family:system-ui,-apple-system,Segoe UI,sans-serif;color:#20312c\">\n  <div style=\"display:flex;gap:10px;align-items:center;flex-wrap:wrap;margin-bottom:15px\">\n    <span style=\"padding:6px 10px;border-radius:999px;background:#dcefe2;color:#286344;font-size:12px;font-weight:800;letter-spacing:.05em;text-transform:uppercase\">MiniMax-reported<\/span>\n    <span style=\"color:#64746d;font-size:13px\">Launch article checked October 26, 2026<\/span>\n  <\/div>\n  <div style=\"overflow-x:auto\">\n    <table style=\"width:100%;min-width:650px;border-collapse:collapse\">\n      <thead><tr><th style=\"padding:11px;text-align:left;background:#e9f3ec;border-bottom:1px solid #cbdccf\">Evaluation or claim<\/th><th style=\"padding:11px;text-align:left;background:#e9f3ec;border-bottom:1px solid #cbdccf\">Published result<\/th><th style=\"padding:11px;text-align:left;background:#e9f3ec;border-bottom:1px solid #cbdccf\">\u305d\u306e\u89e3\u91c8\u65b9\u6cd5<\/th><\/tr><\/thead>\n      <tbody>\n        <tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">BankerToolBench<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\"><strong>76.2<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Tool-oriented work, reported by MiniMax<\/td><\/tr>\n        <tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">DeepSWE v1.1<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\"><strong>73.7%<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Software-engineering benchmark context<\/td><\/tr>\n        <tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">OSWorld V2<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\"><strong>61.7<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Computer-use task performance context<\/td><\/tr>\n        <tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Agents\u2019 Last Exam<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\"><strong>29.6<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Agent evaluation context<\/td><\/tr>\n        <tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">\u30b5\u30a4\u30d0\u30fc\u30b8\u30e0<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\"><strong>82%<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Cybersecurity task context<\/td><\/tr>\n        <tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">\u51fa\u529b\u901f\u5ea6<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\"><strong>200+ TPS peak; 170+ TPS P50<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe5\">Provider-reported throughput claim, not our latency test<\/td><\/tr>\n      <\/tbody>\n    <\/table>\n  <\/div>\n<\/section>\n\n\n\n<figure style=\"margin:26px 0\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/minimax-m31-flash-benchmarks_ebfe452256294e99ad63a342d09150ba.webp\" alt=\"MiniMax M3.1 Flash MiniMax-reported benchmark snapshot\" style=\"display:block;width:100%;height:auto;border-radius:20px\"><figcaption style=\"margin-top:9px;color:#6d7771;font-size:13px\">MiniMax-reported benchmark snapshot; this article does not present it as an independent rerun.<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">These numbers do not tell you the price of one repository fix, the latency of your region, or how well the model follows your tools. For those decisions, use a small controlled evaluation with your own prompts, tool schema, context length, and retry policy.<\/p>\n\n\n\n<h2 id=\"pricing\" class=\"wp-block-heading\">API pricing and cost examples<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">MiniMax\u2019s <a href=\"https:\/\/platform.minimax.io\/docs\/pricing\/overview.md\">pricing overview<\/a> lists USD token rates for the direct API. The billing unit is one million tokens, and cached input is priced separately from uncached input. These rates describe the MiniMax API route; they are not a conversion of GlobalGPT credits or a third-party plan.<\/p>\n\n\n\n<section style=\"box-sizing:border-box;margin:24px 0;padding:22px;border:1px solid #e0d9e4;border-radius:20px;background:#fbf8ff;font-family:system-ui,-apple-system,Segoe UI,sans-serif;color:#282033\">\n  <div style=\"display:grid;grid-template-columns:repeat(auto-fit,minmax(170px,1fr));gap:13px\">\n    <div style=\"padding:17px;border-radius:15px;background:#fff;box-shadow:0 7px 18px #5d477314\"><div style=\"font-size:13px;color:#756c80\">\u30a4\u30f3\u30d7\u30c3\u30c8<\/div><div style=\"margin:5px 0;font-size:30px;font-weight:850\">$0.15<\/div><div style=\"font-size:13px;color:#756c80\">100\u4e07\u30c8\u30fc\u30af\u30f3\u3042\u305f\u308a<\/div><\/div>\n    <div style=\"padding:17px;border-radius:15px;background:#30203f;color:#fff;box-shadow:0 7px 18px #30203f2b\"><div style=\"font-size:13px;color:#e8d8f5\">\u51fa\u529b<\/div><div style=\"margin:5px 0;font-size:30px;font-weight:850\">$0.60<\/div><div style=\"font-size:13px;color:#e8d8f5\">100\u4e07\u30c8\u30fc\u30af\u30f3\u3042\u305f\u308a<\/div><\/div>\n    <div style=\"padding:17px;border-radius:15px;background:#fff;box-shadow:0 7px 18px #5d477314\"><div style=\"font-size:13px;color:#756c80\">\u30ad\u30e3\u30c3\u30b7\u30e5\u3055\u308c\u305f\u5165\u529b<\/div><div style=\"margin:5px 0;font-size:30px;font-weight:850\">$0.015<\/div><div style=\"font-size:13px;color:#756c80\">100\u4e07\u30c8\u30fc\u30af\u30f3\u3042\u305f\u308a<\/div><\/div>\n  <\/div>\n<\/section>\n\n\n\n<div style=\"overflow-x:auto;margin:24px 0\"><table style=\"width:100%;min-width:680px;border-collapse:collapse;font-family:system-ui,-apple-system,Segoe UI,sans-serif\"><thead><tr><th style=\"padding:11px;text-align:left;background:#f2edf7;border-bottom:1px solid #ded4e8\">\u30b7\u30ca\u30ea\u30aa<\/th><th style=\"padding:11px;text-align:left;background:#f2edf7;border-bottom:1px solid #ded4e8\">\u8a08\u7b97<\/th><th style=\"padding:11px;text-align:left;background:#f2edf7;border-bottom:1px solid #ded4e8\">API\u5229\u7528\u6599\u306e\u898b\u7a4d\u3082\u308a<\/th><\/tr><\/thead><tbody><tr><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\">100k input + 20k output<\/td><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\">(0.1 \u00d7 $0.15) + (0.02 \u00d7 $0.60)<\/td><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\"><strong>$0.027<\/strong><\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\">100k cached input + 20k output<\/td><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\">(0.1 \u00d7 $0.015) + (0.02 \u00d7 $0.60)<\/td><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\"><strong>$0.0135<\/strong><\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\">1M input + 200k output<\/td><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\">(1 \u00d7 $0.15) + (0.2 \u00d7 $0.60)<\/td><td style=\"padding:11px;border-bottom:1px solid #ebe5ef\"><strong>$0.27<\/strong><\/td><\/tr><\/tbody><\/table><\/div>\n\n\n\n<p class=\"wp-block-paragraph\">These are arithmetic examples before taxes, account-level discounts, retries, tool-side charges, or any platform markup. The cheapest path is usually to reuse stable prefixes as cacheable input and keep the output budget appropriate to the task instead of paying for a large completion every time.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If you are comparing direct API pricing with hosted subscriptions, keep the units separate. The <a href=\"https:\/\/www.glbgpt.com\/hub\/claude-ai-pricing-2026-the-ultimate-guide-to-plans-api-costs-and-limits\/\">Claude\u306e\u30d7\u30e9\u30f3\u304a\u3088\u3073API\u6599\u91d1\u30ac\u30a4\u30c9<\/a> is a useful example of why consumer plans, API tokens, and platform credits should not be treated as the same currency.<\/p>\n\n\n\n<h2 id=\"api-setup\" class=\"wp-block-heading\">API setup: OpenAI-compatible and Anthropic-compatible<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">MiniMax documents two compatibility layers in its <a href=\"https:\/\/platform.minimax.io\/docs\/api-reference\/text-openai-api.md\">OpenAI-compatible reference<\/a> \u305d\u3057\u3066 <a href=\"https:\/\/platform.minimax.io\/docs\/api-reference\/text-anthropic-api.md\">Anthropic-compatible reference<\/a>. The OpenAI-compatible route is convenient if your application already uses the OpenAI SDK. The Anthropic-compatible route uses Anthropic message semantics. In both cases the documented model ID is <code>MiniMax-M3.1-Flash<\/code>; keep the spelling and capitalization exactly as shown by the target endpoint.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">OpenAI-compatible example<\/h3>\n\n\n\n<div style=\"margin:10px 0\"><button type=\"button\" style=\"padding:8px 12px;border:0;border-radius:8px;background:#557b70;color:#fff;font-weight:700;cursor:pointer\" onclick=\"const c=this.parentElement.nextElementSibling.querySelector('code');navigator.clipboard?.writeText(c.innerText);this.textContent='Copied';setTimeout(()=>this.textContent=&#8217;Copy code&#8217;,1200)&#8221;>Copy code<\/button><\/div><pre style=\"box-sizing:border-box;overflow-x:auto;padding:18px;border-radius:14px;background:#182720;color:#e8f5ec;font:13px1.6 ui-monospace,sfmono-regular,consolas,monospace\"><code>from openai import OpenAI\n\nclient = OpenAI(\n    api_key=\"MINIMAX_API_KEY\",\n    base_url=\"https:\/\/api.minimax.io\/v1\",\n)\n\nresponse = client.chat.completions.create(\n    model=\"MiniMax-M3.1-Flash\",\n    messages=[\n        {\"role\": \"user\", \"content\": \"Summarize this incident and return JSON.\"}\n    ],\n    reasoning_effort=\"medium\",\n    max_completion_tokens=1200,\n)\n\nprint(response.choices[0].message.content)<\/code><\/pre>\n\n\n\n<p class=\"wp-block-paragraph\">For this protocol, MiniMax documents <code>\u63a8\u8ad6\u306e\u52aa\u529b<\/code> \u305d\u3057\u3066 <code>\u6700\u5927\u5b8c\u4e86\u30c8\u30fc\u30af\u30f3\u6570<\/code>. The response can include reasoning metadata, but your application should only display or store it according to your own logging and privacy policy.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Anthropic-compatible example<\/h3>\n\n\n\n<div style=\"margin:10px 0\"><button type=\"button\" style=\"padding:8px 12px;border:0;border-radius:8px;background:#557b70;color:#fff;font-weight:700;cursor:pointer\" onclick=\"const c=this.parentElement.nextElementSibling.querySelector('code');navigator.clipboard?.writeText(c.innerText);this.textContent='Copied';setTimeout(()=>this.textContent=&#8217;Copy code&#8217;,1200)&#8221;>Copy code<\/button><\/div><pre style=\"box-sizing:border-box;overflow-x:auto;padding:18px;border-radius:14px;background:#182720;color:#e8f5ec;font:13px1.6 ui-monospace,sfmono-regular,consolas,monospace\"><code>curl https:\/\/api.minimax.io\/anthropic\/v1\/messages \\\n  -H \"Authorization: Bearer $MINIMAX_API_KEY\" \\\n  -H \"Content-Type: application\/json\" \\\n  -H \"anthropic-version: 2023-06-01\" \\\n  -d '{\n    \"model\": \"MiniMax-M3.1-Flash\",\n    \"max_tokens\": 1200,\n    \"output_config\": {\"effort\": \"medium\"},\n    \"messages\": [\n      {\"role\": \"user\", \"content\": \"Return three repair steps as JSON.\"}\n    ]\n  }'<\/code><\/pre>\n\n\n\n<p class=\"wp-block-paragraph\">Anthropic-compatible calls use <code>output_config.effort<\/code>. The provider\u2019s reference also describes content blocks for text, images, and video, plus tool calls. Check the current protocol reference before copying parameters between the two routes; a parameter name that works in one compatibility layer may be ignored in the other.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For a second perspective on compatible model endpoints, see this <a href=\"https:\/\/www.glbgpt.com\/hub\/claude-api-guide\/\">Claude API\u30ac\u30a4\u30c9<\/a>. It covers a different provider, but the same practical lesson applies: pin the base URL, model ID, protocol, and error handling together.<\/p>\n\n\n\n<h2 id=\"parameters-and-modalities\" class=\"wp-block-heading\">Parameters, thinking, tools, and multimodal limits<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Thinking is forced on for M3.1 Flash according to the official API references. You control its intensity through the documented effort field: <code>\u30ed\u30fc<\/code>, <code>\u4e2d<\/code>, <code>\u9ad8<\/code>, <code>xhigh<\/code>, \u3042\u308b\u3044\u306f <code>max<\/code>. Treat effort as a workload setting rather than a quality guarantee: raise it for difficult planning or debugging, and start lower for short structured transformations.<\/p>\n\n\n\n<div style=\"overflow-x:auto;margin:24px 0\"><table style=\"width:100%;min-width:660px;border-collapse:collapse;font-family:system-ui,-apple-system,Segoe UI,sans-serif\"><thead><tr><th style=\"padding:11px;text-align:left;background:#edf4f1;border-bottom:1px solid #d5e2da\">\u80fd\u529b<\/th><th style=\"padding:11px;text-align:left;background:#edf4f1;border-bottom:1px solid #d5e2da\">Documented detail<\/th><th style=\"padding:11px;text-align:left;background:#edf4f1;border-bottom:1px solid #d5e2da\">Practical implication<\/th><\/tr><\/thead><tbody><tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">\u6587\u8108<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">100\u4e07\u30c8\u30fc\u30af\u30f3<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Large repositories, specs, and logs can fit in one workflow<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">\u8003\u3048\u308b<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Always on; five effort levels<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Tune reasoning depth and budget per task<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">\u30c4\u30fc\u30eb<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Tool calls supported<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Useful for agents that inspect, edit, or verify<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">\u753b\u50cf<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Up to 10 MB<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Resize or compress larger inputs before sending<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Video URL\/base64<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Up to 50 MB<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Use a smaller clip or hosted URL for API requests<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Video through Files API<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Up to 512 MB<\/td><td style=\"padding:11px;border-bottom:1px solid #e4ebe6\">Prefer file upload for larger supported video inputs<\/td><\/tr><\/tbody><\/table><\/div>\n\n\n\n<p class=\"wp-block-paragraph\">Multimodal support does not mean every client library will serialize every content block correctly. Test the exact SDK, content shape, and endpoint you plan to deploy. Keep file size checks on your side so a rejected 50 MB video does not become a confusing generic 4xx response.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Long-context and multimodal systems also make data handling more important. Before sending private repositories, screenshots, or videos, review your retention and access assumptions; this <a href=\"https:\/\/www.glbgpt.com\/hub\/ai-data-privacy\/\">AI\u30c7\u30fc\u30bf\u30d7\u30e9\u30a4\u30d0\u30b7\u30fc\u30ac\u30a4\u30c9<\/a> is a useful checklist for that conversation.<\/p>\n\n\n\n<h2 id=\"hands-on-tests\" class=\"wp-block-heading\">What our GlobalGPT tests returned<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">We tested three real tasks on October 26, 2026: an exact JSON response, a JavaScript debugging fix, and a three-step HTTP 503 diagnosis. Each task completed successfully. The results below are functional observations from that test session, not a general benchmark.<\/p>\n\n\n\n<section style=\"box-sizing:border-box;margin:25px 0;padding:22px;border:1px solid #e7d7d5;border-radius:20px;background:#fff8f6;font-family:system-ui,-apple-system,Segoe UI,sans-serif;color:#322629\">\n  <div style=\"overflow-x:auto\"><table style=\"width:100%;min-width:780px;border-collapse:collapse\"><thead><tr><th style=\"padding:11px;text-align:left;background:#f8e9e6;border-bottom:1px solid #ead2ce\">\u30c6\u30b9\u30c8<\/th><th style=\"padding:11px;text-align:left;background:#f8e9e6;border-bottom:1px solid #ead2ce\">\u30bf\u30b9\u30af<\/th><th style=\"padding:11px;text-align:left;background:#f8e9e6;border-bottom:1px solid #ead2ce\">\u7d50\u679c<\/th><th style=\"padding:11px;text-align:left;background:#f8e9e6;border-bottom:1px solid #ead2ce\">Observed usage<\/th><\/tr><\/thead><tbody><tr><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\"><strong>T1<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">Exact JSON reply<\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\"><code>{\"status\":\"READY\",\"model\":\"MiniMax M3.1 Flash\"}<\/code><\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">5.757s \u00b7 3,958 in \/ 17 out<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\"><strong>T2<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">JavaScript\u306e\u30c7\u30d0\u30c3\u30b0<\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">\u8fd4\u5374\u6e08\u307f <code>Number(a) + Number(b)<\/code> and explained string concatenation<\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">11.976s \u00b7 3,979 in \/ 511 out<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\"><strong>T3<\/strong><\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">Three-step HTTP 503 diagnosis<\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">Identified likely causes and recommended the next diagnostic checks<\/td><td style=\"padding:11px;border-bottom:1px solid #f0e0dd\">11.954s \u00b7 3,991 in \/ 376 out<\/td><\/tr><\/tbody><\/table><\/div>\n  <figure style=\"margin:26px 0\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/minimax-m31-flash-real-tests_11ca40f4e1a34b24b4684afa4a6429fc.webp\" alt=\"Screenshot-style summary of three completed MiniMax M3.1 Flash task runs\" style=\"display:block;width:100%;height:auto;border-radius:20px\"><figcaption style=\"margin-top:9px;color:#6d7771;font-size:13px\">Task-result snapshot from the real-task test: exact JSON, JavaScript debugging, and HTTP 503 diagnosis.<\/figcaption><\/figure>\n  <p style=\"margin:14px 0 0;font-size:13px;line-height:1.6;color:#6e5a6b\">All three generations completed. The platform metadata reported <code>customerChargeAmount: 0<\/code> for these test turns. That is an observation for this authenticated account\/session, not evidence that every GlobalGPT user has free access.<\/p>\n<\/section>\n\n\n\n<figure style=\"margin:26px 0\"><img decoding=\"async\" src=\"https:\/\/static.futureshareai.com\/glb_features\/minimax-m31-flash-real-tests_11ca40f4e1a34b24b4684afa4a6429fc.webp\" alt=\"Three completed MiniMax M3.1 Flash real task checks\" style=\"display:block;width:100%;height:auto;border-radius:20px\"><figcaption style=\"margin-top:9px;color:#6d7771;font-size:13px\">Three real task checks: exact JSON, JavaScript debugging, and HTTP 503 diagnosis.<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">This is a small functional check, not a benchmark. It shows that the verified route handled exact formatting, a short code correction, and a multi-step diagnostic response on the recorded day. It does not establish average latency, throughput, general coding quality, or a stable price across accounts.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For practical use, start with a small request, confirm the model name and request format, then scale after the response and cost match your workflow.<\/p>\n\n\n\n<h2 id=\"free-access\" class=\"wp-block-heading\">Free access, M Plan, and MiniMax Code<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">There are three different questions hidden inside \u201cIs MiniMax M3.1 Flash free?\u201d<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u516c\u5f0fAPI\uff1a<\/strong> the published route has token prices, so it is usage-billed rather than universally free.<\/li>\n\n\n\n<li><strong>MiniMax products:<\/strong> MiniMax says the model is available through the M Plan and MiniMax Code. Plan inclusion, quotas, and regional availability are separate from API billing.<\/li>\n\n\n\n<li><strong>Hosted platforms:<\/strong> our GlobalGPT test account showed a zero customer charge for three completed turns. That observation does not establish a permanent free tier, a public quota, or free access for every account.<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">If you want to explore hosted access without setting up a direct API key, use the <a href=\"https:\/\/www.glbgpt.com\/hub\/all-in-one-ai-models\/\">\u30aa\u30fc\u30eb\u30a4\u30f3\u30ef\u30f3AI\u30e2\u30c7\u30eb\u30ac\u30a4\u30c9<\/a> to compare platform workflows, then verify the current model card and account terms before a larger run. Do not compare a platform\u2019s credits with MiniMax token prices until you know what each unit buys.<\/p>\n\n\n\n<h2 id=\"m31-vs-m3\" class=\"wp-block-heading\">M3.1 Flash vs M3<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">M3.1 Flash is the newer Flash-branded release in the checked October 2026 material. The useful comparison is therefore operational: do you want the newer 1M-context, speed-focused model and its documented thinking controls, or are you maintaining an existing M3 integration that has already been tuned and validated?<\/p>\n\n\n\n<div style=\"overflow-x:auto;margin:24px 0\"><table style=\"width:100%;min-width:650px;border-collapse:collapse;font-family:system-ui,-apple-system,Segoe UI,sans-serif\"><thead><tr><th style=\"padding:11px;text-align:left;background:#edf0f7;border-bottom:1px solid #d9deea\">\u5224\u65ad\u30dd\u30a4\u30f3\u30c8<\/th><th style=\"padding:11px;text-align:left;background:#edf0f7;border-bottom:1px solid #d9deea\">M3.1 Flash<\/th><th style=\"padding:11px;text-align:left;background:#edf0f7;border-bottom:1px solid #d9deea\">Existing M3 workflow<\/th><\/tr><\/thead><tbody><tr><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Context and reasoning<\/td><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">1M context; thinking always on with five effort levels<\/td><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Keep the behavior you have already tested<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Migration risk<\/td><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Validate prompts, tool schemas, JSON contracts, and token budgets<\/td><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Lower immediate change if the current integration is stable<\/td><\/tr><tr><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">\u6b21\u306b\u53d6\u308b\u3079\u304d\u6700\u5584\u306e\u7b56<\/td><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Run a side-by-side sample on your real workload<\/td><td style=\"padding:11px;border-bottom:1px solid #e6e8ef\">Keep as fallback until the new route passes acceptance tests<\/td><\/tr><\/tbody><\/table><\/div>\n\n\n\n<p class=\"wp-block-paragraph\">MiniMax\u2019s checked public material does not give this article a like-for-like M3-versus-M3.1 Flash score table. That is why the comparison above focuses on documented interface and migration decisions instead of inventing a numerical winner.<\/p>\n\n\n\n<h2 id=\"who-should-use-it\" class=\"wp-block-heading\">\u3069\u306e\u3088\u3046\u306a\u4eba\u306b\u9069\u3057\u3066\u3044\u307e\u3059\u304b\uff1f<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>\u30d5\u30a3\u30c3\u30c8\u611f\u826f\u597d\uff1a<\/strong> developers building agents, coding assistants, repository analyzers, computer-use flows, or long-context workflows that benefit from adjustable reasoning effort and tool calls. The low direct input rate also makes it practical for repeated context-heavy experiments when your cache strategy is sound.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Use caution:<\/strong> teams that require a stable, independently verified benchmark, a guaranteed free tier, or a single compatibility endpoint across every provider. Test structured output, tool-call arguments, error handling, and the exact region before moving production traffic.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For a broader evaluation, pair a model test with your actual data-analysis or coding workflow. The <a href=\"https:\/\/www.glbgpt.com\/hub\/best-ai-tools-for-data-analysis-tested\/\">tested AI data-analysis tools guide<\/a> is a useful reminder that a model score alone does not measure the quality of the whole workflow.<\/p>\n\n\n\n<h2 id=\"verdict\" class=\"wp-block-heading\">\u8a55\u6c7a<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">MiniMax M3.1 Flash is a credible low-cost option for agentic coding and long-context work. The direct API price is easy to model, the compatibility layers reduce migration work, and the documented multimodal and tool support make it more useful than a text-only \u201cfast\u201d model. The strongest evidence here is the combination of official interface details and three completed functional tests.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Keep the boundaries clear: MiniMax-reported benchmark scores are not independent measurements, and The observed zero-charge result applies to that test session. Start with a small side-by-side test, then scale only after your prompts, tools, budgets, and access route behave the way your application needs.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If you want a hosted starting point, <a href=\"https:\/\/www.glbgpt.com\/home?inviter=hub_popup&amp;login=1\"><strong>open the GlobalGPT workspace<\/strong><\/a> and check the current MiniMax M3.1 Flash model card before submitting production work.<\/p>\n\n\n\n<h2 id=\"faq\" class=\"wp-block-heading\">\u3088\u304f\u3042\u308b\u3054\u8cea\u554f<\/h2>\n\n\n\n<h3 class=\"wp-block-heading\">What is MiniMax M3.1 Flash?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">MiniMax M3.1 Flash is a 1M-context reasoning model released on October 19, 2026 for agentic coding, computer-use, tool-calling, cybersecurity, and other long-context workflows.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">How much does MiniMax M3.1 Flash cost?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The official API rates checked on October 26, 2026 are $0.15 per million input tokens, $0.60 per million output tokens, and $0.015 per million cached input tokens. These are direct MiniMax API rates.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Does MiniMax M3.1 Flash have a 1M context window?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Yes. MiniMax\u2019s official invocation guide lists a 1M-token context window. Your effective usable context still depends on the endpoint, request shape, tool payloads, and output budget.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Can I call it through the OpenAI SDK?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Yes. MiniMax documents an OpenAI-compatible base URL and the model ID MiniMax-M3.1-Flash. The documented controls include reasoning_effort and max_completion_tokens; validate the exact route before production use.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Can I use an Anthropic-compatible API?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Yes. MiniMax documents an Anthropic-compatible messages route using output_config.effort for thinking effort, along with text, image, video, and tool content blocks.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Is MiniMax M3.1 Flash free?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">There is no universal free API claim in the checked sources. MiniMax lists paid token rates and also names M Plan and MiniMax Code as access routes. The three-turn test session showed zero charge at the time of testing; this observation can vary by plan and account.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">What did the GlobalGPT tests show?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Three real-task runs completed successfully: exact JSON, JavaScript debugging, and a three-step HTTP 503 diagnosis. Observed durations were 5.757, 11.976, and 11.954 seconds; these are single-run observations, not a latency benchmark.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">What are the multimodal input limits?<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The checked API references list images up to 10 MB, video URLs or base64 up to 50 MB, and video through the Files API up to 512 MB. Confirm the current endpoint and client serialization before sending production media.<\/p>\n\n\n\n<script type=\"application\/ld+json\">{\n    \"@context\": \"https:\\\/\\\/schema.org\",\n    \"@type\": \"FAQPage\",\n    \"mainEntity\": [\n        {\n            \"@type\": \"Question\",\n            \"name\": \"What is MiniMax M3.1 Flash?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"MiniMax M3.1 Flash is a 1M-context reasoning model released on October 19, 2026 for agentic coding, computer-use, tool-calling, cybersecurity, and other long-context workflows.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"How much does MiniMax M3.1 Flash cost?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"The official API rates checked on October 26, 2026 are $0.15 per million input tokens, $0.60 per million output tokens, and $0.015 per million cached input tokens. These are direct MiniMax API rates.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Does MiniMax M3.1 Flash have a 1M context window?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Yes. MiniMax\\u2019s official invocation guide lists a 1M-token context window. Your effective usable context still depends on the endpoint, request shape, tool payloads, and output budget.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Can I call it through the OpenAI SDK?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Yes. MiniMax documents an OpenAI-compatible base URL and the model ID MiniMax-M3.1-Flash. The documented controls include reasoning_effort and max_completion_tokens; validate the exact route before production use.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Can I use an Anthropic-compatible API?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Yes. MiniMax documents an Anthropic-compatible messages route using output_config.effort for thinking effort, along with text, image, video, and tool content blocks.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"Is MiniMax M3.1 Flash free?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"There is no universal free API claim in the checked sources. MiniMax lists paid token rates and also names M Plan and MiniMax Code as access routes. The three-turn test session showed zero charge at the time of testing; this observation can vary by plan and account.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"What did the GlobalGPT tests show?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"Three real-task runs completed successfully: exact JSON, JavaScript debugging, and a three-step HTTP 503 diagnosis. Observed durations were 5.757, 11.976, and 11.954 seconds; these are single-run observations, not a latency benchmark.\"\n            }\n        },\n        {\n            \"@type\": \"Question\",\n            \"name\": \"What are the multimodal input limits?\",\n            \"acceptedAnswer\": {\n                \"@type\": \"Answer\",\n                \"text\": \"The checked API references list images up to 10 MB, video URLs or base64 up to 50 MB, and video through the Files API up to 512 MB. Confirm the current endpoint and client serialization before sending production media.\"\n            }\n        }\n    ]\n}<\/script>","protected":false},"excerpt":{"rendered":"<p>Quick answer A fast, low-cost model for long-context agent work MiniMax M3.1 Flash launched on October 19, 2026 with a 1M-token context window, adjustable thinking effort, multimodal input, and tool support. We tested three real tasks\u2014exact JSON, JavaScript debugging, and HTTP 503 diagnosis\u2014and all three completed successfully. Context 1M tokens Input price $0.15 per 1M tokens Real tasks 3\/3 completed If you are comparing low-cost agent models for coding, tool use, or long documents, this guide puts the useful facts in one place: what the model can do, what the official scores actually mean, how to call it, how to estimate cost, and where access is currently visible. You can [&hellip;]<\/p>","protected":false},"author":16,"featured_media":21148,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_seopress_robots_primary_cat":"","_seopress_titles_title":"MiniMax M3.1 Flash Guide: API, Pricing, Benchmarks, Parameters & Free Access","_seopress_titles_desc":"MiniMax M3.1 Flash guide: official benchmarks, $0.15\/M input pricing, API examples, thinking parameters, limits, and free access routes.","_seopress_robots_index":"","footnotes":""},"categories":[7],"tags":[],"class_list":["post-21129","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-chat"],"acf":[],"_links":{"self":[{"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/posts\/21129","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/users\/16"}],"replies":[{"embeddable":true,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/comments?post=21129"}],"version-history":[{"count":5,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/posts\/21129\/revisions"}],"predecessor-version":[{"id":21162,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/posts\/21129\/revisions\/21162"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/media\/21148"}],"wp:attachment":[{"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/media?parent=21129"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/categories?post=21129"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/wp.glbgpt.com\/jp\/wp-json\/wp\/v2\/tags?post=21129"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}