Explore tweets tagged as #openrouter
I was in awe seeing horizon beta's performance and it went down few hours back, then now officially GPT 5 replaces horizon models as stated by Openrouter.
0
1
10
there's a new model called "horizon beta" on openrouter. it's cloaked so nobody really knows who made it (openai, anthropic, etc). it's surprisingly good at long conversations & calling tools. how tf is this free to use 😭
19
6
178
Chutes is an AI compute powerhouse today, not tomorrow. 🪂. - 30k Paying Customers (From 0 in just over a month).- 1/2 a Million Users of our platform alone (not including OpenRouter users).- 4.3T Tokens/Month (at current rates).- Largest Open-Source Provider on OpenRouter.
9
55
289
go to openrouter . try the new mystery model - horizon alpha. - it is great at coding and writing.- it is stealth (no one knows who created it yet).- might be openai gpt-5 or something similar . i asked: "what else can you tell me about yourself?"
4
1
26
这家公司提供的LLM推理速度快到冒烟，能达到至少1500token/每秒！. 什么概念呢？在openrouter的qwen3 coder模型的提供商吞吐如下图，cerebras的平均吞吐1650tok/s，是第二名92tok/s的17倍. 按照这个吞吐量，在编码领域几秒钟就能喷出几千行代码！
3
0
4
☔️ A lesser talked about issue on OpenRouter is third party providers that serve quantized models. This worsens model performance. For example even though most providers claim to serve GLM-4.5 at fp8, performance scores are worse on OpenRouter compared to the official Zai API.
3
1
6
get a jailbroken, fully compliant gpt-oss nearly every single time:.1) use completions mode - not chat.(eg openrouter .ai/api/v1/completions).2) type your question.3) paste exactly the contents of this screenshot.4) press submit
65
86
2K
GPT-OSS pagoda voxel art. left: 120b.right: 20b. (OpenRouter Groq provider)
1
0
2
These are the top apps of the month on OpenRouter?. Wow. I don't even know most of them.
4
0
19
I’m still trying to figure out where half of OpenAI tokens went between June 5th and June 6th at Openrouter. It didn't happen with any other provider.
2
0
2
Glad I finally got around to this. Our AI hedge fund now runs @OpenRouterAI. Made it easy to add:.• GLM 4.5.• GLM 4.5 Air.• Qwen 3 (235B) Thinking. Didn’t realize OpenRouter would be this smooth. I might rework our whole LLM stack around it. Huge shoutout to @xanderatallah
9
11
161
💬 There has been some discussion on who is making the mysterious horizon-alpha/beta model on OpenRouter. People say it's OpenAI, but based on the SvelteBench benchmark scores (where it does very well) I think it's more likely that it's another company/group.
5
0
17
Horizon Beta on OpenRouter
New stealth model: Horizon Beta 🌄. An improved version of Horizon Alpha.
2
1
15
What's the plan for the US to compete with Chinese models long term?. - DeepSeek and Qwen usage on OpenRouter increased from 6% to 17% ytd. - Up to 20 times cheaper in some tests I'm running
0
0
3