[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-96008":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":15,"stars7d":15,"stars30d":16,"stars90d":15,"forks30d":15,"starsTrendScore":15,"compositeScore":17,"rankGlobal":9,"rankLanguage":9,"license":18,"archived":19,"fork":19,"defaultBranch":20,"hasWiki":21,"hasPages":19,"topics":22,"createdAt":9,"pushedAt":9,"updatedAt":23,"readmeContent":24,"aiSummary":25,"trendingCount":15,"starSnapshotCount":15,"syncStatus":13,"lastSyncTime":26,"discoverSource":27},96008,"real-api-pricing","FeiZhuLulu\u002Freal-api-pricing","FeiZhuLulu","真实 API 定价：订阅月费 ÷ 实际可用 token，含额度、单价与分榜帕累托图。Real API pricing: subscription fee ÷ usable tokens, with allowance, unit-price, and leaderboard Pareto charts.",null,"HTML",286,11,2,8,0,174,53.24,"MIT License",false,"main",true,[],"2026-09-20 04:01:32","**English** | [中文](README.zh.md)\n\n# Real API Pricing\n\n**Real unit price = monthly subscription fee ÷ monthly usable tokens.**\n\nFull adopted data is shown first, followed by one Pareto chart per leaderboard. Monthly figures default to four weeks of saturated use; vendor-defined monthly pools remain as defined (Kimi's monthly pool is 5× its weekly pool). Input, output and cache tokens are all included. Prices use a logarithmic axis, with cheaper points farther right.\n\nDollar\u002Fcredit pools and three-part token prices are converted with one project-wide standard workload: **97.5% cache reads, 2.15% fresh input, and 0.35% output**. This is a comparison convention, not a claim about any provider's actual workload. Measurements that already report total tokens—dashboard back-calculations, local usage logs, controlled saturation tests, and official absolute-token tables—are not normalized again. Where only total tokens and a cost-weighted percentage are available but the token-type split is unknown, the observed total is retained and the limitation is recorded rather than inventing a split. Cache writes are not modeled separately; where a provider charges for them, converted token allowances may be overstated. See [conventions](data\u002Fconventions.json) and the [token-mix audit](data\u002Fresearch\u002Ftoken-mix-audit-round2-2026-09-07.json).\n\nGLM Coding Plan is recomputed from Zhipu's official weekly credits and cache\u002Finput\u002Foutput coefficients under the same standard workload. Peak, midpoint and off-peak scenarios are shown separately instead of copying the official 95%-cache example table. A Caijing saturation-cost test and community evidence are consistent in scale, but there is still no fully specified independent V3 Pro\u002FMax saturation test. See the [official-table archive](data\u002Fresearch\u002Fquotas-web-2026-09.json) and [community-evidence review](data\u002Fresearch\u002Fglm-community-round1-2026-09-07.json).\n\nEach chart uses scores from its named leaderboard only. Code Arena here specifically means the WebDev Overall Arena Score, not general coding ability. GPT-5.6 Luna now uses a ChatGPT Plus dashboard measurement: 112.67 million total tokens consumed about 6% of the weekly allowance, giving 7.511 billion tokens\u002Fmonth for Plus. The 5x and 20x plans are scaled from that measured Plus baseline, so the rightmost Luna point is 150.222 billion tokens\u002Fmonth at medium confidence rather than the superseded 240.24 billion Sol-credit derivation. Claude Max's 15.7 billion-token estimate applies to the permanent terms from September 14, 2026, not a promotional ceiling. Chinese charts use 100-million-token units: 77.37 in Chinese equals 7.737 billion in English.\n\n**[All charts: English \u002F 中文, SVG \u002F PNG](charts\u002FREADME.md)** · [English files](charts\u002Fen\u002F) · [中文文件](charts\u002Fzh\u002F)\n\n## Data snapshot\n\nSnapshot: 2026-09-07. Each row is one **plan × actual served model**; allowances of different models under the same plan are alternatives and must not be added together.\n\n| Coverage | Rows |\n|---|---:|\n| All adopted plan × model points | 184 |\n| Subscription points with monthly allowance | 173 |\n| Metered API baselines | 11 |\n| OpenCode Go \u002F Command Code GOAT \u002F Ollama models | 28 \u002F 38 \u002F 20 |\n| Code Arena \u002F Agent Arena scored points | 134 \u002F 138 |\n| AA Intelligence \u002F AA Coding Agent scored points | 131 \u002F 59 |\n\n**Download the data:** [adopted values (CSV)](data\u002Fadopted.csv) · [computed points (CSV)](derived\u002Fpoints.csv) · [computed points (JSON)](derived\u002Fpoints.json) · [data notes and score coverage](data\u002FREADME.md) · [dated evidence](data\u002Fresearch\u002F)\n\n## Monthly allowance overview\n\nAll 173 subscription plan × model points, sorted by monthly usable tokens. The standard chart keeps a single logarithmic scale; the hybrid-scale view makes the two very large ChatGPT allowances easier to compare.\n\n[English SVG](charts\u002Fen\u002Foverview\u002Fmonthly-allowance-overview.svg) · [中文 SVG](charts\u002Fzh\u002Foverview\u002F额度总览.svg) · [English PNG](charts\u002Fen\u002Foverview\u002Fmonthly-allowance-overview.png) · [中文 PNG](charts\u002Fzh\u002Foverview\u002F额度总览.png) · [Hybrid-scale view](charts\u002Fen\u002Foverview\u002Fmonthly-allowance-overview-hybrid-scale.svg)\n\n![Monthly allowance overview](charts\u002Fen\u002Foverview\u002Fmonthly-allowance-overview.svg)\n\n**Full table:** [English TXT](charts\u002Fen\u002Foverview\u002Fmonthly-allowance-overview-table.txt) · [中文 TXT](charts\u002Fzh\u002Foverview\u002F额度总览表.txt)\n\n## Real unit price overview\n\nAll 184 subscription and API points on one comparable $\u002FMTok scale.\n\n[English SVG](charts\u002Fen\u002Foverview\u002Freal-price-overview.svg) · [中文 SVG](charts\u002Fzh\u002Foverview\u002F单价总览.svg) · [English PNG](charts\u002Fen\u002Foverview\u002Freal-price-overview.png) · [中文 PNG](charts\u002Fzh\u002Foverview\u002F单价总览.png)\n\n![Real unit price overview](charts\u002Fen\u002Foverview\u002Freal-price-overview.svg)\n\n**Full table:** [English TXT](charts\u002Fen\u002Foverview\u002Freal-price-overview-table.txt) · [中文 TXT](charts\u002Fzh\u002Foverview\u002F单价总览表.txt)\n\n## Pareto charts by leaderboard\n\nUsing Real API Pricing as a new baseline, we plot each leaderboard's scores on the Y-axis to redraw its Pareto frontier; the connected line represents that frontier. Subscriptions and metered APIs follow the same dominance rule and both participate in frontier selection.\n\n### Code Arena\n\n[English SVG](charts\u002Fen\u002Fpareto\u002Fpareto-code-arena.svg) · [中文 SVG](charts\u002Fzh\u002Fpareto\u002F帕累托_CodeArena榜.svg) · [English PNG](charts\u002Fen\u002Fpareto\u002Fpareto-code-arena.png) · [中文 PNG](charts\u002Fzh\u002Fpareto\u002F帕累托_CodeArena榜.png)\n\n![Code Arena](charts\u002Fen\u002Fpareto\u002Fpareto-code-arena.svg)\n\n### Agent Arena\n\n[English SVG](charts\u002Fen\u002Fpareto\u002Fpareto-agent-arena.svg) · [中文 SVG](charts\u002Fzh\u002Fpareto\u002F帕累托_AgentArena榜.svg) · [English PNG](charts\u002Fen\u002Fpareto\u002Fpareto-agent-arena.png) · [中文 PNG](charts\u002Fzh\u002Fpareto\u002F帕累托_AgentArena榜.png)\n\n![Agent Arena](charts\u002Fen\u002Fpareto\u002Fpareto-agent-arena.svg)\n\n### AA Intelligence\n\n[English SVG](charts\u002Fen\u002Fpareto\u002Fpareto-aa-intelligence.svg) · [中文 SVG](charts\u002Fzh\u002Fpareto\u002F帕累托_AA智力榜.svg) · [English PNG](charts\u002Fen\u002Fpareto\u002Fpareto-aa-intelligence.png) · [中文 PNG](charts\u002Fzh\u002Fpareto\u002F帕累托_AA智力榜.png)\n\n![AA Intelligence](charts\u002Fen\u002Fpareto\u002Fpareto-aa-intelligence.svg)\n\n### AA Coding Agent\n\n[English SVG](charts\u002Fen\u002Fpareto\u002Fpareto-aa-coding-agent.svg) · [中文 SVG](charts\u002Fzh\u002Fpareto\u002F帕累托_AA编程Agent榜.svg) · [English PNG](charts\u002Fen\u002Fpareto\u002Fpareto-aa-coding-agent.png) · [中文 PNG](charts\u002Fzh\u002Fpareto\u002F帕累托_AA编程Agent榜.png)\n\n![AA Coding Agent](charts\u002Fen\u002Fpareto\u002Fpareto-aa-coding-agent.svg)\n\nAA Coding Agent scores describe tested harness × model × effort configurations. Static charts and `points.*` are explicitly **highest archived configuration reference summaries**. They are not measurements of each subscription\u002FAPI channel; quota-measurement effort and product harness alignment remain unverified. Higher effort does not automatically change $\u002FMTok; it can change tokens consumed per task.\n\n[All-configuration interactive view (Chinese)](charts\u002Fzh\u002Fpareto\u002F帕累托交互图.html) defaults to every archived configuration and offers the highest-score summary as an option. Download the HTML and open it locally with network access for Plotly. All configurations currently use reference mappings, not a verified product-configuration frontier.\n\nThe [configuration archive (JSON)](derived\u002Fbenchmark-configurations.json) \u002F [CSV](derived\u002Fbenchmark-configurations.csv) retains all 128 records, original labels, known harness\u002Feffort, 30 source score intervals, and 70 source task-cost records. The [plan-to-configuration mappings (JSON)](derived\u002Fbenchmark-points.json) \u002F [CSV](derived\u002Fbenchmark-points.csv) contains 604 explicit references, including lower-effort variants. Composer Standard\u002FFast require their own mode; a missing mode stays unscored. Unknown harnesses, efforts and intervals stay null.\n\nSource mean and median task costs are separate fields, not subscription task costs. Score intervals are preserved and available in interactive hover details, but uncertainty does not yet change frontier membership. Numerical quota ranges, robust-frontier analysis and workload sensitivity remain follow-up work; qualitative confidence labels are not numerical error bars.\n\n## Method and reproduction\n\n[Build instructions](BUILD.md) · [Data documentation](data\u002FREADME.md) · [Sources and attribution](SOURCES.md)\n\n## License and acknowledgements\n\nOriginal software: [MIT](LICENSE). Data references include [Awesome Coding Plan](https:\u002F\u002Fgithub.com\u002Fmahonzhan\u002Fawesome-coding-plan) (CC BY 4.0) and the Caijing article 《Token经济，中国账本》. See [SOURCES.md](SOURCES.md) for attribution, changes and third-party terms.\n","该项目提供真实 API 定价分析工具，通过‘月订阅费 ÷ 实际可用 token 数’计算可比单位价格，而非仅依赖厂商标称配额。核心功能包括：统一转换不同计费模式（美元池、信用池、三类 token 分计）为等效 token 用量，采用标准化工作负载（97.5% 缓存读\u002F2.15% 新输入\u002F0.35% 输出）进行横向对比；生成按榜单（如 Code Arena、GPT-5.6 Luna）划分的帕累托图，支持中英文双语 SVG\u002FPNG 图表输出。适用于 AI 服务采购评估、大模型 API 成本效益分析及多厂商定价策略比对场景。","2026-09-08 02:30:07","CREATED_QUERY"]