OpenRouter

Introducing Pareto Code: a new, free, experimental coding router Set `min_coding_score` in your request and route to the cheapest code-capable model that clears your bar, ranked by @ArtificialAnlys. See the Pareto frontier shifting in real time👇

last year I started building an "AI Council" app of different models but decided it was not worth the effort. This evening, I asked @claudeai to rebuild this idea (1 prompt) with @OpenRouter to get some diverse AI perspective on something, and its just...working perfectly

We now allow you to bring your @OpenRouter key into Clark - our state of the art digital labor agent. By bringing your key you get SOTA agent - #1 on GDPval-AA while running lower tier models. You also have @e2b sandbox for each of your agent executions. Each agent is equipped with file system, full browser and hyper optimized to complete complex tasks on cloud computer. Clark can help you find 1000 candidates for a job. Or analyze 100s of documents in parallel. Or make a small cute video game. Think Manus on steroids, and you can choose the model you want to run!

OpenRouter is now supported in @warpdotdev! ❤️

@just_andydev Thanks, fixing!

@OpenRouter Your doc is out dated

@_dr5w Here's the price: openrouter.ai/deepseek/deeps…

@OpenRouter Because it's free Just tell us what the #1 free model is and then we will know this chart too :D

DeepSeek V4 Flash has topped the weekly leaderboard

@XCSme @AIBenchy No, they do not

@OpenRouter @AIBenchy Thanks! I do see this, the LLMs automatically added this: I'll go with it, but I was afraid that adding this info might let LLMs know when they are tested/train on tests. Do the LLM providers get this information (which app has sent each request)?

Huh? How does @OpenRouter know my requests come from @AIBenchy testing? It's cool to see it, but I never created that app on OpenRouter, and don't remember adding any identifier to my requests.

Learn how to use explicit caching with Qwen models: #alibaba-qwen" target="_blank" rel="nofollow noopener">openrouter.ai/docs/guides/be…

Read more about the model:

The new Qwen3.7-Max from @Alibaba_Qwen is live on OpenRouter. The flagship of the Qwen3.7 series, built for agent-centric work: coding, office and productivity tasks, and long-horizon autonomous execution. Big jumps in coding and agent benchmarks over Qwen3.6, with explicit prompt caching for repeated context.

@plane_tg correct!

@OpenRouter hmm so if i keep spamming requests the router doesnt flip me mid conversation? big if true for long convos

TIP 💡 You don't have to worry about cache misses for the Auto Router (in addition to all individual models) OpenRouter will keep your session pinned to one model/provider until your cache expires

05.19.2026 GMI Cloud hits ~90B tokens of DeepSeek V4 in one day on @OpenRouter , ~290B in one week. Thank you to everyone who built with us 🙌

Read more about the model: x.com/GoogleDeepMind…

Try it now: openrouter.ai/google/gemini-…

Gemini 3.5 Flash from @GoogleDeepMind is live on OpenRouter! Beats Gemini 3.1 Pro on coding, agentic work, and tool use at Flash-tier price and speed. 1M context, 65K max output, multimodal. $1.50/M input, $9/M output.

Parallel Search is now in @OpenRouter. Use Parallel for configurable context size, domain filtering, and clean, structured results that deliver the best accuracy and token efficiency.

Migrating from the old web plugin: - Before: "plugins": [{ "id": "web" }] - After: "tools": [{ "type": "openrouter:web_search" }] Try it in the chatroom or read the docs. Full migration guide in the post: openrouter.ai/announcements/…

Both tools support allowed_domains and blocked_domains when you pick a non-native engine. Critical if you're building an agent that should only touch your docs or trusted sources.

Any tool-calling model on OpenRouter can call web search and web fetch agentically. The model decides when to search, what to search for, and how many times. We've added @p0 as a new web search provider. Learn more: openrouter.ai/announcements/…