Codex 与 Mirror

Name: Coding Agent Index
Creator: Artificial Analysis
License: https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf

根据 Artificial Analysis Coding Agent Index比较 Codex 与 Mirror，包括基准测试得分、成本、执行时间和 token 用量。

有关我们方法论的详细信息，请参阅方法论页面。

浏览其他比较

Coding Agent Index

Artificial Analysis Coding Agent Index v2.0 · Higher is better

每个任务的时间

Average agent wall time per task · Lower is better

每个任务的成本

Average API cost per task (USD) · Lower is better

比较

Codex 与 Mirror 的并排比较。

编程智能体比较

指标	Codex GPT-6 Astra (Max)	Mirror Beam 1.0	分析
智能体 Harness	Codex	Mirror
代表模型	GPT-6 Astra (Max)	Beam 1.0
Coding Agent Index	55	20	Codex 的 Coding Agent Index 高于 Mirror
FrontierCode v1.1	49%	20%	Codex 的 FrontierCode v1.1 分数高于 Mirror
Terminal-Bench 4.0	56%	4%	Codex 的 Terminal-Bench 4.0 分数高于 Mirror
SWE-Atlas-QnA-AA	59%	38%	Codex 的 SWE-Atlas-QnA-AA 分数高于 Mirror
每个任务的成本	US$7.04	US$0.32	Mirror 的每个任务成本低于 Codex
每个任务的时间	24.8分钟	38.7分钟	Codex 的每个任务时间短于 Mirror
每个任务的轮次	39.4	93	Codex 的每个任务轮次少于 Mirror
每个任务的 Token 用量	3.23M	4.89M	Codex 的每个任务 token 用量少于 Mirror
缓存命中率	93%	95%	Mirror 的缓存命中率高于 Codex

模型变体

Codex 与 Mirror 的已评估模型变体。

模型变体


Codex	GPT-6 Astra (Max)	max	55	49%	56%	59%	US$7.04	24.8分钟	3.23M
Codex	GPT-6.1 Sol (Max)	max	54	49%	53%	60%	US$1.50	23.7分钟	3.94M
Codex	GPT-6.1 Terra (Xhigh)	xhigh	48	45%	46%	54%	US$0.69	16.4分钟	6.89M
Codex	GPT-6.1 Terra (Max)	max	48	42%	48%	54%	US$1.24	26.1分钟	12.3M
Codex	DeepSeek V4.1 Flash (Max)	max	47	53%	28%	61%	US$0.39	38.1分钟	27.2M
Codex	GPT-6.1 Terra (High)	high	46	43%	42%	53%	US$0.54	12.7分钟	5.6M
Codex	GPT-6.1 Terra (Medium)	medium	42	44%	35%	47%	US$0.39	10.2分钟	4.08M
Codex	GPT-6.1 Terra (Non-reasoning)	non-reasoning	39	46%	25%	46%	US$0.45	9.5分钟	5.78M
Codex	GPT-6.1 Terra (Low)	low	35	41%	25%	39%	US$0.20	6.3分钟	1.89M
Codex	GPT-6 Luna (Max)	max	31	40%	16%	38%	US$0.16	21.3分钟	8.89M
Mirror	Beam 1.0	reasoning	20	20%	4%	38%	US$0.32	38.7分钟	4.89M

性能

Artificial Analysis Coding Agent Index上的性能。

Artificial Analysis Coding Agent Index

Artificial Analysis Coding Agent Index v2.0 incorporates 3 benchmarks: FrontierCode v1.1, Terminal-Bench 4.0, and SWE-Atlas-QnA-AA · Higher is better

着色方式

The Artificial Analysis Coding Agent Index is a composite score built from FrontierCode v1.1、Terminal-Bench 4.0和SWE-Atlas-QnA-AA.

It is useful for quick comparison, but it should be read alongside the per-eval breakdowns. Two agents with similar index values can still have different strengths across repository tasks, terminal workflows, and rubric-based evaluations.

Artificial Analysis Coding Agent Index vs. 每个任务的成本

Artificial Analysis Coding Agent Index vs. 每个任务的平均按 token 计费 API 成本（USD）

着色方式

Most attractive quadrant

Pareto line

Each point represents a coding-agent variant. Farther left means lower average cost per task, while higher on the chart means higher benchmark performance. The most efficient agents sit toward the upper-left: stronger results at lower cost.

Token 用量

Artificial Analysis Coding Agent Index上的 token 消耗。

每个任务的 Token 用量

每个任务的平均输入、缓存和输出 token

提示词缓存命中率可能因提供商路由而显著变化，从而实质性地影响实际成本。

Non-cached input tokens sent to the model, including prompts, instructions, tool context, and task context that were not served from prompt cache.

Artificial Analysis Coding Agent Index vs. 总 Token

Artificial Analysis Coding Agent Index vs. 每个任务的平均总 Token

着色方式

Most attractive quadrant

Each point represents a coding-agent variant. Farther left means lower average total token usage per task, while higher on the chart means higher benchmark performance. Agents toward the upper-left achieve stronger results with fewer tokens.

成本

基于当前按 token 计费价格的 Artificial Analysis Coding Agent Index按 token 计费 API 成本。

每个任务的成本

Average pay-per-token API cost per task (USD) · Lower is better

着色方式

This chart shows the average pay-per-token API cost per task across the Artificial Analysis Coding Agent Index, spanning FrontierCode v1.1、Terminal-Bench 4.0和SWE-Atlas-QnA-AA.

Where applicable, that cost model includes standard input pricing, discounted cached-input pricing, separate cache-write charges, and output pricing rather than treating all prompt tokens as if they were billed at the same uncached input rate.

It is intended to show pay-per-token API cost, not consumer plan pricing or the full operational cost of deploying the system in production. Infrastructure, engineering, and supervision costs are not the focus of this metric.

执行时间

Artificial Analysis Coding Agent Index上的智能体活跃运行时间。

每个任务的时间

Average agent wall time per task · Lower is better

着色方式

This chart uses agent wall time: how long the agent process was actively running on each task.

It does not include environment startup, verifier or judge time, or other harness overhead, so it is a cleaner comparison of how long the agent itself was working.

Artificial Analysis Coding Agent Index vs. 执行时间

Artificial Analysis Coding Agent Index vs. 每个任务的平均智能体运行时间

着色方式

Most attractive quadrant

Each point represents a coding-agent variant. Farther left means shorter average agent runtime per task, while higher on the chart means higher benchmark performance. Agents toward the upper-left deliver stronger results in less active agent time.

Codex 与 Mirror

Coding Agent Index

每个任务的时间

每个任务的成本

比较

编程智能体比较

模型变体

模型变体

性能

Artificial Analysis Coding Agent Index

What This Metric Means

Artificial Analysis Coding Agent Index vs. 每个任务的成本

How to Read This Chart

Token 用量

每个任务的 Token 用量

Input Tokens

Artificial Analysis Coding Agent Index vs. 总 Token

How to Read This Chart

成本

每个任务的成本

What Cost Is Measuring

执行时间

每个任务的时间

What Execution Time Is Measuring

Artificial Analysis Coding Agent Index vs. 执行时间

How to Read This Chart