惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

L
Lohrmann on Cybersecurity
Martin Fowler
Martin Fowler
Engineering at Meta
Engineering at Meta
腾讯CDC
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
Microsoft Azure Blog
Microsoft Azure Blog
G
Google Developers Blog
TaoSecurity Blog
TaoSecurity Blog
博客园_首页
Vercel News
Vercel News
Hugging Face - Blog
Hugging Face - Blog
D
Darknet – Hacking Tools, Hacker News & Cyber Security
Last Week in AI
Last Week in AI
cs.CL updates on arXiv.org
cs.CL updates on arXiv.org
T
The Exploit Database - CXSecurity.com
量子位
Project Zero
Project Zero
A
Arctic Wolf
小众软件
小众软件
NISL@THU
NISL@THU
C
CERT Recently Published Vulnerability Notes
有赞技术团队
有赞技术团队
MongoDB | Blog
MongoDB | Blog
博客园 - 聂微东
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
N
News and Events Feed by Topic
宝玉的分享
宝玉的分享
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
T
Troy Hunt's Blog
P
Privacy & Cybersecurity Law Blog
Security Latest
Security Latest
B
Blog
酷 壳 – CoolShell
酷 壳 – CoolShell
D
DataBreaches.Net
Schneier on Security
Schneier on Security
The Hacker News
The Hacker News
K
Kaspersky official blog
C
Check Point Blog
Hacker News: Ask HN
Hacker News: Ask HN
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
Webroot Blog
Webroot Blog
www.infosecurity-magazine.com
www.infosecurity-magazine.com
人人都是产品经理
人人都是产品经理
AI
AI
Cisco Talos Blog
Cisco Talos Blog
MyScale Blog
MyScale Blog
Cloudbric
Cloudbric
B
Blog RSS Feed
S
Schneier on Security
P
Palo Alto Networks Blog

OpenRouter Blog

Choosing the Optimal Image Input Detail Level in LLMs — OpenRouter Blog DeepSeek V4 Is Earning Agentic Token Share — OpenRouter Blog The Open Weight Models that Matter: June 2026 — OpenRouter Blog The OpenRouter MCP Server — OpenRouter Blog Introducing the Unified Image API — OpenRouter Blog The AI Governance Checklist That Maps to Your Stack — OpenRouter Blog Enforce AI Data Residency at the Routing Layer — OpenRouter Blog OpenRouter vs Portkey: Routing Network vs Control Plane — OpenRouter Blog OpenRouter vs LiteLLM: Managed vs Self-Hosted Gateway — OpenRouter Blog Connect OpenClaw to OpenRouter: One Key, Failover, Free Models — OpenRouter Blog Connect SillyTavern to OpenRouter: Setup, Models, Fixes — OpenRouter Blog A Robot is Sprinting Towards You: Do You Want it Running on Claude or Grok? Codex CLI with OpenRouter: config.toml Setup and Models — OpenRouter Blog Claude Code with OpenRouter: Setup, Models, and Costs — OpenRouter Blog How to Use OpenRouter With Any Coding Agent or AI Tool — OpenRouter Blog Subagent: Let Your Model Delegate the Busywork — OpenRouter Blog Free LLM API in 2026: 13 Options Ranked and Compared — OpenRouter Blog How to Enforce Agentic AI Governance at the API Layer — OpenRouter Blog Keep Your Agent Running When Models Disappear — OpenRouter Blog Hermes Agent + OpenRouter: Setup, Model Choice & Routing Config — OpenRouter Blog Lowest-Cost LLM Inference: The Complete OpenRouter Guide — OpenRouter Blog How OpenRouter Model Routing Works: Providers, Fallbacks & Auto Router — OpenRouter Blog OpenRouter Failover: Provider Failover vs Model Fallbacks Explained — OpenRouter Blog Surpassing Frontier Performance with Fusion — OpenRouter Blog Dinner is Served — OpenRouter Blog LLM Gateway: What It Is and How to Choose One — OpenRouter Blog Advisor: Give Any Model a Lifeline to a Smarter One — OpenRouter Blog Gemini 2.5 Flash API - Pricing, Quickstart & Provider Comparison — OpenRouter Blog EU AI Act & Colorado ADMT Compliance: Human Oversight for AI Agents — OpenRouter Blog May Release Spotlight — OpenRouter Blog Guardrails: Protect your Agents, Data, and Costs — OpenRouter Blog OpenRouter Raises $113M Series B — OpenRouter Blog Human-in-the-Loop Tools for the Agent SDK — OpenRouter Blog Consistent Web Search and Fetch Across Every Model — OpenRouter Blog GPT-5.5 Price Increase: What It Actually Costs — OpenRouter Blog New Audio APIs for Speech and Transcription — OpenRouter Blog Response Caching: Zero Cost for Identical Requests — OpenRouter Blog April Release Spotlight — OpenRouter Blog Create OpenRouter Accounts via CLI with Stripe Projects — OpenRouter Blog Opus 4.7 Agent SDK: Building Multi-turn Agent Workflows on OpenRouter — OpenRouter Blog Build Your Own Harness with the Agent SDK — OpenRouter Blog Introducing Workspaces — OpenRouter Blog Announcing Video Generation — OpenRouter Blog Auto Exacto: Adaptive Quality Routing, On by Default — OpenRouter Blog February Release Spotlight — OpenRouter Blog OpenRouter Outages on February 17 and 19, 2026 — OpenRouter Blog January Release Spotlight — OpenRouter Blog Distillable Models and Synthetic Data Pipelines with NeMo Data Designer — OpenRouter Blog December Release Spotlight — OpenRouter Blog Response Healing: Reduce JSON Defects by 80%+ — OpenRouter Blog The 2025 State of AI Report — OpenRouter Blog Is Implicit Caching Prompt Retention? — OpenRouter Blog Provider Variance: Introducing Exacto — OpenRouter Blog 1 million free BYOK requests per month — OpenRouter Blog The First-Ever Image Model Is Up on OpenRouter — OpenRouter Blog GPT-5 is now live — OpenRouter Blog Audio Inputs and PDF URLs for Apps — OpenRouter Blog Presets: How To Seamlessly Transfer Model Configurations Across Apps — OpenRouter Blog New Privacy-Focused Provider Drop: Venice — OpenRouter Blog Use OpenRouter Models in Cursor: Try it with Moonshot AI Updates to Our Free Tier: Sustaining Accessible AI for Everyone — OpenRouter Blog New Stealth Model: "Cypher Alpha" — OpenRouter Blog Introducing Presets: Manage LLM Configs from Your Dashboard! — OpenRouter Blog Dev & BYOK Updates: Uptime API + Smarter Key Management — OpenRouter Blog Simplifying Our Platform Fee — OpenRouter Blog GIF Prompts, Omni Search, Tool Caching, and BYOK Flags — OpenRouter Blog New Features: Reasoning Streams, Crypto Invoices, End-User IDs & More — OpenRouter Blog Passkeys, DevEx Upgrades, and a New Guide for TypeScript Agents — OpenRouter Blog New Provider Drop: Cerebras Is Here — OpenRouter Blog Better Insights, Faster Metrics, and New Developer Power Tools — OpenRouter Blog Privacy Clarity, New Providers, OAuth Upgrade, and Gemini Gets Parallel Tools — OpenRouter Blog Universal PDF Support — OpenRouter Blog Smarter Charts, Inline SVGs, and Live Usage Accounting — OpenRouter Blog Quasar Alpha and Optimus Alpha Reveal — OpenRouter Blog "Stealth" model: Optimus Alpha — OpenRouter Blog “Stealth” model: Quasar Alpha — OpenRouter Blog Never Pay for Empty AI Responses Again — OpenRouter Blog Deep Research & Many New Models — OpenRouter Blog Introducing Nitro and Floor Price Shortcuts — OpenRouter Blog Introducing Cloudflare as a new provider — OpenRouter Blog Reasoning Tokens for Thinking Models — OpenRouter Blog Introducing Web Search via the API — OpenRouter Blog Standardized finish reasons — OpenRouter Blog Happy New Year! Introducing a new Auto Router — OpenRouter Blog Holiday launches: Web Search & Price Cuts — OpenRouter Blog Bring Your Own API Keys — OpenRouter Blog Crypto Payments API — OpenRouter Blog Structured Outputs & Free Gemini Flash 2.0 — OpenRouter Blog Price Drops and Llama 3.3 70b — OpenRouter Blog Author Pages & Amazon Nova — OpenRouter Blog
Kilo Code + OpenRouter: Setup, Routing, and Free Models — OpenRouter Blog
OpenRouter · 2026-06-18 · via OpenRouter Blog

Kilo Code runs the coding session: chat, edits, context, modes, and the model picker inside your editor or terminal. It doesn’t ship its own models. It expects you to bring a provider.

OpenRouter is that provider layer. Add it and Kilo gets one API key for 300+ models across 70+ providers, plus routing rules, failover, and billing in one place. Kilo also offers Kilo Gateway, its own managed-billing path, which we compare below. The whole setup is three steps, and the routing controls live in one config file.

Add OpenRouter to Kilo Code in three steps

First, create a key on your API Keys page. Sign in with Google or GitHub. Load credits on the pricing page unless you’re sticking to free models. There’s no minimum spend on pay-as-you-go.

Second, open Settings in Kilo Code from the gear icon, go to the Providers tab, click Connect on OpenRouter, paste the key, and save.

Third, open the model picker and type a name or slug to switch models. Browse the live list at openrouter.ai/models, since the catalog changes as providers ship new models.

Kilo writes that provider setup to kilo.json (or ~/.config/kilo/kilo.json), so the VS Code extension and the CLI share one config. The Providers tab edits the file for you; CLI users can edit it directly. A saved key with no credits still fails on the first paid request, so load credits before testing a paid model.

Control routing and long prompts from kilo.json

Kilo forwards anything under a model’s options straight to OpenRouter as providerOptions.openrouter, without validating it. That pass-through is what gives you OpenRouter’s provider routing inside Kilo. Put the provider block under models.{modelId}.options:

{
  "provider": {
    "openrouter": {
      "models": {
        "anthropic/claude-sonnet-4.5": {
          "options": {
            "provider": {
              "sort": "price",
              "order": ["Anthropic", "Google"],
              "only": ["Anthropic"],
              "data_collection": "deny",
              "zdr": true
            }
          }
        }
      }
    }
  }
}

Each field maps to an OpenRouter routing parameter: sort by price, throughput, or latency; order to set provider preference; only to fail rather than leave an approved list; data_collection and zdr for data policy. The nesting has to be exact. A provider block at the root provider.openrouter level does nothing, and because Kilo passes fields through unvalidated, a misspelled field is ignored with no error. Match the path above exactly.

For prompts that run past a model’s context window, OpenRouter’s middle-out transform drops content from the middle, where models pay the least attention, instead of rejecting the request. Turn it on in the same options block:

{ "options": { "transforms": ["middle-out"] } }

Run free models inside Kilo

OpenRouter has 20+ free models. Filter for them by typing free in Kilo’s model picker. With no credits loaded, free models run at 50 requests per day and 20 per minute. Add $10 or more in credits and that daily cap rises to 1,000.

Keep the two free paths straight. OpenRouter free models use your OpenRouter account and rate limits. Kilo’s kilo-auto/free, served through Kilo Gateway, doesn’t touch your OpenRouter credits or routing rules.

One caveat for private code: free endpoints can log prompts, so they’re a poor fit for proprietary source unless the data policy works for you. Add data_collection: "deny" to narrow eligible providers.

Choose OpenRouter or Kilo Gateway

Both put models in front of Kilo Code, so the choice comes down to what your team already runs.

Reach for OpenRouter when you already use it across tools, want provider routing controls, or want one account and one bill spanning every AI tool you run. It charges 5.5% on credit purchases with no provider markup, and BYOK lets you route through your own provider key for 5% of what the provider would bill, waived for the first 1M requests each month.

Kilo Gateway is the native path: Kilo markets it as zero-markup, billing lands in your Kilo account, and routing is Kilo-managed. If Kilo Code is your only AI surface and it carries the exact model you need, compare that model in Kilo Gateway before taking on the OpenRouter fee. It mostly comes down to account gravity. If OpenRouter already sits in your stack, add it to Kilo and keep one routing policy.

Frequently asked questions

Is OpenRouter free with Kilo Code?

Partly. OpenRouter has 20+ free models you can use inside Kilo Code, capped at 50 requests per day and 20 per minute with no credits loaded. Add $10 or more in credits and the free-model cap rises to 1,000 requests per day. Kilo’s own kilo-auto/free option is separate and runs through Kilo Gateway.

What is the difference between Kilo Code and OpenRouter?

Kilo Code is the coding agent that runs in your editor or terminal. OpenRouter is a provider you add inside Kilo Code to reach 300+ models from 70+ providers through one API key. You use them together.

How do I set the OpenRouter provider order in Kilo Code?

Add an order array under models.{modelId}.options.provider in kilo.json, listing provider names in preference order such as ["Anthropic", "Google"]. Add only to restrict requests to an approved list. Kilo forwards these fields to OpenRouter without validating them.

Does OpenRouter cost more than Kilo’s own gateway?

OpenRouter charges a 5.5% fee on credit purchases and doesn’t mark up provider prices. Kilo markets Kilo Gateway as zero-markup. OpenRouter is often the better operational choice when you already use it across tools or need provider routing controls.

Can I use OpenRouter with the Kilo CLI?

Yes. The CLI reads the same kilo.json config, so you can add the OpenRouter provider block and routing options directly. The VS Code Providers tab writes to the same file.