huggingface-best

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

HuggingFace Best Model Finder

HuggingFace最佳模型查找工具

Finds the best models for a task by querying official HF benchmark leaderboards, enriching results with model size data, filtering for what fits on the user's device, and returning a comparison table with benchmark scores.

通过查询官方HF基准排行榜，结合模型规模数据丰富结果，根据用户设备进行筛选，并返回带有基准分数的对比表格，为任务找到最佳模型。

Step 1: Parse the request

步骤1：解析请求

Extract from the user's message:

Task: what they want the model to do (coding, math/reasoning, chat, OCR, RAG/retrieval, speech recognition, image classification, multimodal, agents, etc.)
Device: hardware constraints (MacBook M-series 8/16/32/64GB unified memory, RTX GPU with VRAM amount, CPU-only, cloud/no constraint, etc.)

If device is not mentioned, skip filtering entirely and return the highest-performing models regardless of size. If the task is genuinely ambiguous, ask one clarifying question.

从用户消息中提取：

任务：用户希望模型完成的工作（编码、数学/推理、聊天、OCR、RAG/检索、语音识别、图像分类、多模态、Agent等）
设备：硬件限制（MacBook M系列8/16/32/64GB统一内存、带特定VRAM容量的RTX GPU、仅CPU、云端/无限制等）

如果未提及设备，则完全跳过筛选，返回性能最高的模型，无论其规模如何。如果任务确实不明确，提出一个澄清问题。

Device → max parameter budget

设备 → 最大参数预算

When a device is specified, extract its available memory (unified RAM for Apple Silicon, VRAM for discrete GPUs) and apply:

fp16 max params (B) ≈ memory (GB) ÷ 2
Q4 max params (B) ≈ memory (GB) × 2

Examples: 16GB → 8B fp16 / 32B Q4 — 24GB VRAM → 12B fp16 / 48B Q4 — 8GB → 4B fp16 / 16B Q4

当指定设备时，提取其可用内存（Apple Silicon的统一RAM、独立GPU的VRAM）并应用以下规则：

fp16最大参数（B） ≈ 内存（GB）÷ 2
Q4最大参数（B） ≈ 内存（GB）× 2

示例：16GB → 8B fp16 / 32B Q4 — 24GB VRAM → 12B fp16 / 48B Q4 — 8GB → 4B fp16 / 16B Q4

Step 2: Find relevant benchmark datasets

步骤2：查找相关基准数据集

Fetch the full list of official HF benchmarks:

bash

curl -s -H "Authorization: Bearer $(cat ~/.cache/huggingface/token)" \
  "https://huggingface.co/api/datasets?filter=benchmark:official&limit=500" | jq '[.[] | {id, tags, description}]'

Read the returned list and select the datasets most relevant to the user's task — match on dataset id, tags, and description. Use your judgment; don't limit yourself to 2-3. Aim for comprehensive coverage: if 5 benchmarks clearly cover the task, use all 5.

获取官方HF基准的完整列表：

bash

curl -s -H "Authorization: Bearer $(cat ~/.cache/huggingface/token)" \
  "https://huggingface.co/api/datasets?filter=benchmark:official&limit=500" | jq '[.[] | {id, tags, description}]'

读取返回的列表，选择与用户任务最相关的数据集——根据数据集ID、标签和描述进行匹配。请自行判断，不要局限于2-3个。力求全面覆盖：如果有5个基准明确涵盖该任务，则全部使用。

Step 3: Fetch top models from leaderboards

步骤3：从排行榜获取顶级模型

For each selected benchmark dataset:

bash

curl -s -H "Authorization: Bearer $(cat ~/.cache/huggingface/token)" \
  "https://huggingface.co/api/datasets/<namespace>/<repo>/leaderboard" | jq '[.[:15] | .[] | {rank, modelId, value, verified}]'

Collect model IDs and scores across all benchmarks. If a leaderboard returns an error (404, 401, etc.), skip it and note it in the output.

对于每个选定的基准数据集：

bash

curl -s -H "Authorization: Bearer $(cat ~/.cache/huggingface/token)" \
  "https://huggingface.co/api/datasets/<namespace>/<repo>/leaderboard" | jq '[.[:15] | .[] | {rank, modelId, value, verified}]'

收集所有基准中的模型ID和分数。如果排行榜返回错误（404、401等），则跳过该排行榜并在输出中注明。

Step 4: Enrich with model metadata

步骤4：丰富模型元数据

For the top 10-15 candidate model IDs, get model infos.

bash

undefined

针对前10-15个候选模型ID，获取模型信息。

bash

undefined

REST API

curl -s -H "Authorization: Bearer $(cat ~/.cache/huggingface/token)"
"https://huggingface.co/api/models/org/model1" | jq '{safetensors, tags, cardData}'

CLI (hf-cli)

hf models info org/model1 --json | jq '{safetensors, tags, cardData}'


Extract from each response:
- **Parameters**: `safetensors.total` → convert to B (e.g., 7_241_748_480 → "7.2B")
- **License**: from model card tags (look for `license:apache-2.0`, `license:mit`, etc.)
- If `safetensors` is absent, parse size from the model name (look for "7b", "8b", "13b", "70b", "72b", etc.)

---

hf models info org/model1 --json | jq '{safetensors, tags, cardData}'


从每个响应中提取：
- **参数**：`safetensors.total` → 转换为B（例如，7_241_748_480 → "7.2B"）
- **许可证**：从模型卡片标签中获取（查找`license:apache-2.0`、`license:mit`等）
- 如果`safetensors`不存在，从模型名称中解析规模（查找"7b"、"8b"、"13b"、"70b"、"72b"等）

---

Step 5: Filter and rank

步骤5：筛选和排序

If a device was specified:

Remove models exceeding the fp16 parameter budget for the device
Flag models that fit only with Q4 quantization (multiply budget by ~4 for Q4 capacity)
If a highly-ranked model is slightly over budget, keep it with a "needs Q4" note — don't silently drop it

If no device was mentioned: skip all size filtering — just rank by benchmark score.

Then: rank by benchmark score (descending), keep top 5-8 models.

Include proprietary models (GPT-4, Claude, Gemini) if they appear on leaderboards, but flag them as "API only / not self-hostable". If the user explicitly asked for local/open models only, exclude them.

如果指定了设备：

移除超出设备fp16参数预算的模型
标记仅通过Q4量化才能适配的模型（将预算乘以约4得到Q4容量）
如果某个排名靠前的模型略微超出预算，保留它并添加“需要Q4量化”的注释——不要直接删除

如果未提及设备： 跳过所有规模筛选——仅按基准分数排序。

然后：按基准分数降序排序，保留前5-8个模型。

如果专有模型（GPT-4、Claude、Gemini）出现在排行榜中，将其纳入，但标记为“仅API可用 / 无法自托管”。如果用户明确要求仅本地/开源模型，则排除它们。

Step 6: Output

步骤6：输出

Comparison table

对比表格

markdown

| # | Model | Params | [Benchmark 1] | [Benchmark 2] | License | On device |
|---|-------|--------|--------------|--------------|---------|-----------|
| ⭐1 | [org/name](https://huggingface.co/org/name) | 7B | 85.2% | — | Apache 2.0 | Yes (fp16) |
| 2 | [org/name](https://huggingface.co/org/name) | 13B | 83.1% | 71.5% | MIT | Q4 only |
| 3 | [org/name](https://huggingface.co/org/name) | 70B | 90.0% | 81.0% | Llama | Too large |

Link model names to
```
https://huggingface.co/<model_id>
```
Use
```
—
```
for benchmarks where the model wasn't evaluated
Star the top recommended pick with ⭐
"On device" values:
```
Yes (fp16)
```
,
```
Q4 only
```
,
```
Too large
```
,
```
API only
```

markdown

| # | Model | Params | [Benchmark 1] | [Benchmark 2] | License | On device |
|---|-------|--------|--------------|--------------|---------|-----------|
| ⭐1 | [org/name](https://huggingface.co/org/name) | 7B | 85.2% | — | Apache 2.0 | Yes (fp16) |
| 2 | [org/name](https://huggingface.co/org/name) | 13B | 83.1% | 71.5% | MIT | Q4 only |
| 3 | [org/name](https://huggingface.co/org/name) | 70B | 90.0% | 81.0% | Llama | Too large |

将模型名称链接到
```
https://huggingface.co/<model_id>
```
对于未评估该模型的基准，使用
```
—
```
用⭐标记排名第一的推荐模型
“On device”取值：
```
Yes (fp16)
```
、
```
Q4 only
```
、
```
Too large
```
、
```
API only
```

Follow-up

后续跟进

After presenting the table, ask the user: "Would you like to run [top recommended model]?"

If they say yes, ask whether they'd prefer to:

Run locally — ask about their device if not already known, then give appropriate setup instructions
Run on HF Jobs — point them to the HF Jobs guide: https://huggingface.co/docs/huggingface_hub/en/guides/jobs

展示表格后，询问用户：“是否要运行**[排名第一的推荐模型]**？”

如果用户同意，询问他们偏好：

本地运行 — 如果尚未了解设备信息，询问设备详情，然后提供相应的设置说明
在HF Jobs上运行 — 引导他们查看HF Jobs指南：https://huggingface.co/docs/huggingface_hub/en/guides/jobs

Error handling

错误处理

Leaderboard not found: skip, note "leaderboard unavailable" in output
Model missing from hub_repo_details: fall back to parsing size from model name
No benchmarks found for task: use the curated fallback table above, or try
```
hub_repo_search
```
with
```
filters=["<task>"]
```
sorted by
```
trendingScore
```
All leaderboards fail: fall back to
```
hub_repo_search
```
for popular models tagged with the task, note that results are by popularity rather than benchmark score

未找到排行榜：跳过，在输出中注明“排行榜不可用”
模型在hub_repo_details中缺失：退而求其次，从模型名称解析规模
未找到适用于该任务的基准：使用上述精选的备用表格，或尝试使用
```
hub_repo_search
```
并设置
```
filters=["<task>"]
```
，按
```
trendingScore
```
排序
所有排行榜均失败：退而求其次，使用
```
hub_repo_search
```
查找带有该任务标签的热门模型，并注明结果是按受欢迎程度而非基准分数排序