发现模型

基于 OpenRouter 真实使用情况、生产环境性能和独立基准测试的实用起点。

浏览全部 624 模型
今日前沿

Claude Fable 5.2 在前沿模型中领先

Anthropic 在 Artificial Analysis 智能指数上以 5 分的优势略微领先于 Google。

模型智能水平(0–100)

Auto is OpenRouter’s flagship model router, built for general-purpose usage. It adapts to each request, choosing cost-optimal models and adjusting reasoning effort. It balances task difficulty, benchmark results, live prices, and prompt caching. Use it for apps, chat, and general agents. For agent sessions, it follows progress across turns, keeps the same model while work is going well, and adjusts reasoning effort or switches models when the task gets stuck. It respects all your governance settings, including guardrails, ZDR, and budgets.

Auto Code is OpenRouter’s router built specifically for coding agents and harnesses. It uses the same routing engine as the [Auto Router](https://dynamicsinfo.com/openrouter/auto), with every request treated as part of an agent session and model choices guided by coding-task pass rate, completion time, and cost per solved task. It follows edits, test results, and recurring errors across turns, keeping the model while work is going well and trying more reasoning effort before switching when trouble persists. Its default coding tier starts from stronger models and weighs failed code changes more heavily. Your model and provider restrictions still apply.

The experimental version of our Auto Router where we test new improvements. Use it to get the latest and greatest version of our general purpose auto router, but expect beta quality. When new improvements are proven in auto-beta, we graduate them to the auto router. We've love to hear your feedback on whether auto-beta is working for you, or requests for routing improvements. To see which model was used, visit [Activity](/activity), or read the `model` attribute of the response. Your response will be priced at the same rate as the routed model. Learn more, including how to customize the models for routing, in our [docs](/docs/guides/routing/routers/auto-router).

Fusion turns your prompt into a small multi-model deliberation. A panel of expert models (see below) analyzes your prompt in parallel with web search and web fetch enabled, then a judge model synthesizes their responses into a structured analysis — consensus, contradictions, partial coverage, unique insights, and blind spots — and writes the final answer from it. By default the panel is the `Quality` preset; switch to `Budget` for cheaper members, or override the panel and judge entirely with the fusion plugin's `analysis_models` and `model` fields. Reach for Fusion when a single model isn't enough — research, expert critique, or anywhere the cost of being wrong outweighs a few extra completions. Because Fusion runs every panel member plus a judge call, your request is priced as the sum of those underlying completions rather than a single model. To see which models ran, visit [Activity](/activity). Note: Context limits are dependent on which models you select. Learn more in our [docs](/docs/guides/features/plugins/fusion). For another way to route, see the [Auto Router](/openrouter/auto).

The Pareto Router maintains a tiered shortlist of strong coding models, ranked by [Artificial Analysis](https://artificialanalysis.ai/) coding percentiles. Set min_coding_score between 0 and 1 on the [pareto-router plugin](https://dynamicsinfo.com/docs/guides/routing/routers/pareto-router#the-min_coding_score-parameter) to control how strong a coder you need; higher scores select stronger (and typically more expensive) models. If you omit min_coding_score, the router defaults to the High tier. Selecting Nitro from the variant dropdown ranks the models in your tier by measured throughput and routes each request to the fastest one, so you trade some model variety for lower latency. Read the [Pareto Router docs](https://dynamicsinfo.com/docs/guides/routing/routers/pareto-router) for the full selection logic, fallback behavior, and how to customize routing. For another way to route, see the [Auto Router](/openrouter/auto).

性价比领先模型

按标价计算,每美元可获得的能力最多。

排名模型智能水平百分位编程百分位智能体能力百分位token
1
z-ai 的网站图标
GLM 5.3 Flash
40799111.9T
2
deepseek 的网站图标
DeepSeek V4 Flash 0731 (batch)
–757382.5M
3
inclusionai 的网站图标
Ling 3.0 Flash VL
–60576B
4
anthropic 的网站图标
Claude Haiku 5.5 (batch)
65––30.4B
5
deepseek 的网站图标
DeepSeek V4 Flash 0423
–59542.2T

速度最快的模型

在多个提供商中输出速度中位数最高。

排名模型吞吐量TTFT p50提供商
1
qwen 的网站图标
Qwen3.8 27B
883 tok/s253ms23
2
openai 的网站图标
gpt-oss-120b
832 tok/s138ms26
3
deepseek 的网站图标
DeepSeek V4.1 Flash 1017
492 tok/s209ms48
4
meta-llama 的网站图标
Llama 3.1 8B Instruct
470 tok/s143ms4
5
openai 的网站图标
gpt-oss-20b
423 tok/s117ms12

始终指向最新版本的别名

固定一个模型 ID,即可始终获取该系列的最新版本。

排名模型系列 token上下文
1
perplexity 的网站图标
Decider Flash Latest
32.5B262K
2
anthropic 的网站图标
Claude Fable Latest
1T1M
3
google 的网站图标
Gemini Flash Latest
4.1T1M
4
z-ai 的网站图标
GLM Latest
5.1T1M
5
liquid 的网站图标
d1-ultra Latest
44.1B33K

免费模型

使用最多且提供免费端点的模型。

排名模型token/周上下文
1
nvidia 的网站图标
Nemotron 3 Ultra (free)
3.2T1M
2
nvidia 的网站图标
Nemotron 3.5 Super VL (free)
3T1M
3
baidu 的网站图标
ERNIE-5.5-Flash-Preview (free)
1.9T1M
4
nvidia 的网站图标
Nemotron 3.5 Lightning (free)
1.2T1M
5
thinkingmachines 的网站图标
Inkling 2 Small (free)
1.1T1M

图像模型

通过领先的模型提供商生成和编辑图像。

查看全部

视频模型

通过领先的模型提供商,根据文本和图像提示生成视频。

查看全部

探索每种模态

从一组专为您的任务精选的模型开始。

浏览所有模型在模型页面搜索、筛选、排序和比较完整目录中的模型。
查看全部 624 个模型