Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/tako-dev/cli/model-benchmarknpx skills add tako-dev/cli --skill model-benchmarkgit clone --depth 1 https://github.com/tako-dev/cliWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00075 | $0.01307 |
| Opus 5 | $0.00037 | $0.00654 |
| Sonnet 5 | $0.00015 | $0.00261 |
| Haiku 4.5 | $0.00007 | $0.00131 |
Grade A, and why
model-benchmark scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tako 模型选择指南 (benchmark 2026-06-10)
通过 Tako Provider API 评测 16 个模型在 7 个编码任务上的表现。 下表合并了能力评分与场景推荐，按推荐度排序——直接选第一个符合你场景的模型即可。
选哪个模型（能力 + 推荐合并）
| 模型 | 综合分 | 速度 | 最适合场景 | 备注 |
|---|---|---|---|---|
| mimo-v2.5-pro | 7/7 | 中 | 最强编码 / 复杂任务 / 拿不准时的默认选择 | 唯一全通过 |
| claude-sonnet-4-6 | 6.5/7 | 中 | 通用最佳 / 复杂推理 / 长上下文(1M) | 综合最稳 |
| minimax-m2.5 | 6.5/7 | 快(3-10s) | 性价比首选 / 批量任务 | 快且准 |
| deepseek-v4-flash | 6/7 | 最快(2-8s) | 快速简单任务 / 高并发初筛 | 简单任务全对 |
| gpt-5.5 / gpt-5.4 | 6/7 | 中 | OpenAI 生态 / Codex 后端 | — |
| claude-opus-4-6 | 6/7 | 慢 | 追求最高质量、不在意成本时 | — |
| claude-opus-4-8 | 5.5/7 | 慢 | 复杂任务 + thinking | 最新 opus |
| deepseek-v4-pro / mimo-v2.5 | 5.5/7 | 中 | 国产替代、成本敏感 | — |
| glm-5 / claude-opus-4-7 | 5/7 | 中 | 备选 | — |
评分为 7 个编码任务通过数。其余未列模型（minimax-m3/deepseek-3.2/glm-5.1/qwen3.7-max） 因 API 不稳定（502/504）评分不可信，暂不推荐。
详细评测分项
| Model | Fib | Bug Fix | Explain | Refactor | Types | Observer | Constraints | Score |
|---|---|---|---|---|---|---|---|---|
| mimo-v2.5-pro | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | 7/7 |
| claude-sonnet-4-6 | ✓ | ✓ | ◐ | ✓ | ✓ | ✓ | ✓ | 6.5/7 |
| minimax-m2.5 | ✓ | ✓ | ✓ | ✓ | ✓ | ◐ | ✓ | 6.5/7 |
| deepseek-v4-flash | ✓ | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | 6/7 |
| gpt-5.4 | ✓ | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | 6/7 |
| gpt-5.5 | ✓ | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | 6/7 |
| claude-opus-4-6 | ✓ | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | 6/7 |
| claude-opus-4-8 | ✓ | ✓ | ◐ | ✓ | ✓ | ✗ | ✓ | 5.5/7 |
| deepseek-v4-pro | ✓ | ✓ | ✓ | ✓ | ◐ | ✗ | ✓ | 5.5/7 |
| mimo-v2.5 | ✓ | ✓ | ✓ | ✓ | ◐ | ✗ | ✓ | 5.5/7 |
| glm-5 | ✓ | ✓ | ✓ | ✓ | ✓ | ✗ | ⚠ | 5/7 |
| claude-opus-4-7 | ◐ | ✓ | ◐ | ✓ | ✓ | ✗ | ✓ | 5/7 |
| minimax-m3 | ✓ | ✓ | ✓ | ⚠ | ⚠ | ✗ | ✓ | 4/7* |
| deepseek-3.2 | ⚠ | ✓ | ⚠ | ✓ | ⚠ | ✗ | ✓ | 3/7* |
| glm-5.1 | ✓ | ✓ | ⚠ | ⚠ | ⚠ | ⚠ | ✓ | 3/7* |
| qwen3.7-max | ⚠ | ⚠ | ⚠ | ⚠ | ⚠ | ⚠ | ⚠ | 0/7* |
*⚠ = API 不可用（502/504），非模型能力问题
测试任务说明
| ID | 难度 | 任务 |
|---|---|---|
| TP-BENCH-01 | Easy | 写 fibonacci 函数 (Python) |
| TP-BENCH-02 | Easy | 修复 off-by-one bug |
| TP-BENCH-03 | Easy | 解释 debounce 代码 |
| TP-BENCH-04 | Medium | 函数拆分重构 |
| TP-BENCH-05 | Medium | TypeScript 泛型类型补充 |
| TP-BENCH-06 | Hard | 实现 EventEmitter (观察者模式 + 泛型) |
| TP-BENCH-07 | Hard | 多约束代码生成 |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 69 lines · 75 tokens per session scan A 137debb57219
model-benchmark is a skill published in the GitHub repository tako-dev/cli (3 stars, last pushed 2d ago), licensed MIT. It adds 75 tokens to every session and 1,307 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…