Claude 3.7 Sonnet Thinking shows strength across a variety of categories
arena 发布的一张图背后的数字,用原生图表重画。对照原图,研究数据,再做你自己的。
简化重绘
由我们用下面的数字重新绘制。悬停、点按或用方向键查看。这是简化重绘:它用我们自己的样式展示数字,不是原图的样子。强调色标出这张图的主角。表格里带 ~ 的是估算值。
数字
150 个点中前 10 个免费| 分组 | 系列 | 数值 | 来源 | 置信 |
|---|---|---|---|---|
| gpt-4.5-preview-2025-02-27 | Overall | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Overall | ~3 | 估算 | 0.50 |
| grok-3-preview-02-24 | Overall | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Overall | ~3 | 估算 | 0.50 |
| o1-2024-12-17 | Overall | ~6 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Overall | ~14 | 估算 | 0.50 |
| deepseek-r1 | Overall | ~6 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Overall | ~3 | 估算 | 0.50 |
| o1-preview | Overall | ~9 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Overall | ~16 | 估算 | 0.50 |
Max 数据,目前留邮箱免费看
查看全部 150 个数据点并下载文件
这张图的完整表格、JSON、CSV 和 agent skill 文件。它们属于 Max;原型阶段,留邮箱即可解锁。
| gemini-2.0-flash-001 | Overall | ~6 | 估算 | 0.50 |
|---|---|---|---|---|
| gemma-3-27b-it | Overall | ~8 | 估算 | 0.50 |
| qwen2.5-max | Overall | ~9 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Overall | ~27 | 估算 | 0.50 |
| deepseek-v3 | Overall | ~12 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Overall w/ Style Control | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Overall w/ Style Control | ~2 | 估算 | 0.50 |
| grok-3-preview-02-24 | Overall w/ Style Control | ~3 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Overall w/ Style Control | ~3 | 估算 | 0.50 |
| o1-2024-12-17 | Overall w/ Style Control | ~3 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Overall w/ Style Control | ~4 | 估算 | 0.50 |
| deepseek-r1 | Overall w/ Style Control | ~4 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Overall w/ Style Control | ~6 | 估算 | 0.50 |
| o1-preview | Overall w/ Style Control | ~7 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Overall w/ Style Control | ~7 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Overall w/ Style Control | ~11 | 估算 | 0.50 |
| gemma-3-27b-it | Overall w/ Style Control | ~11 | 估算 | 0.50 |
| qwen2.5-max | Overall w/ Style Control | ~11 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Overall w/ Style Control | ~11 | 估算 | 0.50 |
| deepseek-v3 | Overall w/ Style Control | ~12 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Hard Prompts | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Hard Prompts | ~5 | 估算 | 0.50 |
| grok-3-preview-02-24 | Hard Prompts | ~5 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Hard Prompts | ~3 | 估算 | 0.50 |
| o1-2024-12-17 | Hard Prompts | ~3 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Hard Prompts | ~5 | 估算 | 0.50 |
| deepseek-r1 | Hard Prompts | ~4 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Hard Prompts | ~3 | 估算 | 0.50 |
| o1-preview | Hard Prompts | ~6 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Hard Prompts | ~15 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Hard Prompts | ~5 | 估算 | 0.50 |
| gemma-3-27b-it | Hard Prompts | ~13 | 估算 | 0.50 |
| qwen2.5-max | Hard Prompts | ~5 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Hard Prompts | ~15 | 估算 | 0.50 |
| deepseek-v3 | Hard Prompts | ~15 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Hard Prompts w/ Style Control | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Hard Prompts w/ Style Control | ~3 | 估算 | 0.50 |
| grok-3-preview-02-24 | Hard Prompts w/ Style Control | ~3 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Hard Prompts w/ Style Control | ~2 | 估算 | 0.50 |
| o1-2024-12-17 | Hard Prompts w/ Style Control | ~2 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Hard Prompts w/ Style Control | ~1 | 估算 | 0.50 |
| deepseek-r1 | Hard Prompts w/ Style Control | ~3 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Hard Prompts w/ Style Control | ~5 | 估算 | 0.50 |
| o1-preview | Hard Prompts w/ Style Control | ~5 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Hard Prompts w/ Style Control | ~11 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Hard Prompts w/ Style Control | ~2 | 估算 | 0.50 |
| gemma-3-27b-it | Hard Prompts w/ Style Control | ~15 | 估算 | 0.50 |
| qwen2.5-max | Hard Prompts w/ Style Control | ~7 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Hard Prompts w/ Style Control | ~15 | 估算 | 0.50 |
| deepseek-v3 | Hard Prompts w/ Style Control | ~16 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Coding | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Coding | ~1 | 估算 | 0.50 |
| grok-3-preview-02-24 | Coding | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Coding | ~3 | 估算 | 0.50 |
| o1-2024-12-17 | Coding | ~3 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Coding | ~7 | 估算 | 0.50 |
| deepseek-r1 | Coding | ~1 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Coding | ~3 | 估算 | 0.50 |
| o1-preview | Coding | ~4 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Coding | ~8 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Coding | ~4 | 估算 | 0.50 |
| gemma-3-27b-it | Coding | ~15 | 估算 | 0.50 |
| qwen2.5-max | Coding | ~6 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Coding | ~7 | 估算 | 0.50 |
| deepseek-v3 | Coding | ~14 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Math | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Math | ~14 | 估算 | 0.50 |
| grok-3-preview-02-24 | Math | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Math | ~2 | 估算 | 0.50 |
| o1-2024-12-17 | Math | ~1 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Math | ~1 | 估算 | 0.50 |
| deepseek-r1 | Math | ~3 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Math | ~2 | 估算 | 0.50 |
| o1-preview | Math | ~2 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Math | ~12 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Math | ~2 | 估算 | 0.50 |
| gemma-3-27b-it | Math | ~6 | 估算 | 0.50 |
| qwen2.5-max | Math | ~3 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Math | ~14 | 估算 | 0.50 |
| deepseek-v3 | Math | ~14 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Creative Writing | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Creative Writing | ~1 | 估算 | 0.50 |
| grok-3-preview-02-24 | Creative Writing | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Creative Writing | ~1 | 估算 | 0.50 |
| o1-2024-12-17 | Creative Writing | ~6 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Creative Writing | ~6 | 估算 | 0.50 |
| deepseek-r1 | Creative Writing | ~2 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Creative Writing | ~1 | 估算 | 0.50 |
| o1-preview | Creative Writing | ~9 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Creative Writing | ~6 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Creative Writing | ~2 | 估算 | 0.50 |
| gemma-3-27b-it | Creative Writing | ~7 | 估算 | 0.50 |
| qwen2.5-max | Creative Writing | ~4 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Creative Writing | ~18 | 估算 | 0.50 |
| deepseek-v3 | Creative Writing | ~17 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Instruction Following | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Instruction Following | ~3 | 估算 | 0.50 |
| grok-3-preview-02-24 | Instruction Following | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Instruction Following | ~2 | 估算 | 0.50 |
| o1-2024-12-17 | Instruction Following | ~3 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Instruction Following | ~3 | 估算 | 0.50 |
| deepseek-r1 | Instruction Following | ~6 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Instruction Following | ~2 | 估算 | 0.50 |
| o1-preview | Instruction Following | ~8 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Instruction Following | ~11 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Instruction Following | ~6 | 估算 | 0.50 |
| gemma-3-27b-it | Instruction Following | ~8 | 估算 | 0.50 |
| qwen2.5-max | Instruction Following | ~8 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Instruction Following | ~7 | 估算 | 0.50 |
| deepseek-v3 | Instruction Following | ~13 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Longer Query | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Longer Query | ~1 | 估算 | 0.50 |
| grok-3-preview-02-24 | Longer Query | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Longer Query | ~1 | 估算 | 0.50 |
| o1-2024-12-17 | Longer Query | ~3 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Longer Query | ~3 | 估算 | 0.50 |
| deepseek-r1 | Longer Query | ~4 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Longer Query | ~2 | 估算 | 0.50 |
| o1-preview | Longer Query | ~7 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Longer Query | ~6 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Longer Query | ~7 | 估算 | 0.50 |
| gemma-3-27b-it | Longer Query | ~7 | 估算 | 0.50 |
| qwen2.5-max | Longer Query | ~6 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Longer Query | ~6 | 估算 | 0.50 |
| deepseek-v3 | Longer Query | ~7 | 估算 | 0.50 |
| gpt-4.5-preview-2025-02-27 | Multi-Turn | ~1 | 估算 | 0.50 |
| chatgpt-4o-latest-20250129 | Multi-Turn | ~2 | 估算 | 0.50 |
| grok-3-preview-02-24 | Multi-Turn | ~1 | 估算 | 0.50 |
| gemini-2.0-pro-exp-02-05 | Multi-Turn | ~2 | 估算 | 0.50 |
| o1-2024-12-17 | Multi-Turn | ~7 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219-thinking-32k | Multi-Turn | ~9 | 估算 | 0.50 |
| deepseek-r1 | Multi-Turn | ~2 | 估算 | 0.50 |
| gemini-2.0-flash-thinking-exp-01-21 | Multi-Turn | ~4 | 估算 | 0.50 |
| o1-preview | Multi-Turn | ~7 | 估算 | 0.50 |
| claude-3-7-sonnet-20250219 | Multi-Turn | ~7 | 估算 | 0.50 |
| gemini-2.0-flash-001 | Multi-Turn | ~5 | 估算 | 0.50 |
| gemma-3-27b-it | Multi-Turn | ~7 | 估算 | 0.50 |
| qwen2.5-max | Multi-Turn | ~8 | 估算 | 0.50 |
| claude-3-5-sonnet-20241022 | Multi-Turn | ~11 | 估算 | 0.50 |
| deepseek-v3 | Multi-Turn | ~8 | 估算 | 0.50 |
Max 订阅者可以通过 API 取同样的数据。 API 文档
API · Max
用 API 调用这张图
Max 套餐可以取这张图的数字(每个数字都带来源),格式为 JSON 或 CSV,附重画规格;也可以让智能体通过 MCP 调用。图表 id:ad22ff385c58。
| 路径 | 返回 | 套餐 | |
|---|---|---|---|
| GET | /api/v1/charts/ad22ff385c58 | 标题、发布者、日期、图型、数据点数;用 Max 密钥时附数据链接 | 开放 |
| GET | /api/v1/charts/ad22ff385c58/data | 每个数据点及其来源和置信度,加上重画规格。加 ?format=csv 得到 CSV | Max |
| GET | /api/v1/charts/ad22ff385c58/render?view=1 | 本页使用的只读绘图视图 | 开放 |
| POST | /mcp | 工具 get_chart_data 加这个 id,供 Claude、Cursor 等智能体使用 | Max |
curl -s https://viral-chart.com/api/v1/charts/ad22ff385c58/data \
-H "Authorization: Bearer $VIRALCHART_KEY"claude mcp add --transport http viralchart https://viral-chart.com/mcp \
--header "Authorization: Bearer $VIRALCHART_KEY"每个密钥每分钟 60 次,超出返回 429 和 Retry-After。每个账号最多 5 个密钥,在账户页创建,只显示一次。没有 Max 时数据接口返回 402,并附套餐链接。这些数字描述的是发布者拥有的图,每个文件都带授权说明。
做一张这样的图
这张图的形式,你的数字。
实时预览。所有内容都留在你的设备上。
选择形式
简化形式:用我们自己的样式,不是原图的样子。你的数字从这张图的数据开始。
换上你的数字
这是这张图自己的名称和数字,方便你看清形式。直接改写任何名称或数字,预览会随之重画。发布前请把每一个都换成你的。
这是这张图的数字,方便你看清形式。发布前请把每一个都换成你的。
你的品牌
导出图片
免费图片的角落有一行小字 “Made with ViralChart”,别人就是通过这一行找到我们的。
Anthropic 的更多数据
全部图表数据常见问题
这些数字有多准?
它们由两遍独立读图估算得出,置信度 0.50(0 到 1)。图上印着的数字读得准,从坐标轴读出的值可能有几个百分点的偏差。
这些数据可以直接用吗?
这些数字描述的是 arena 发布的图表。图片、版式和样式归他们所有。使用数字时请注明发布者,图片请用你自己的数据来做。
怎么拿到文件?
JSON、CSV 和 agent skill 文件在本页的邮箱解锁区。原型里它们是静态文件:/api/charts/ad22ff385c58.json 和 /api/charts/ad22ff385c58.skill.md。