Refuseless

To apologize for the issues our API had yesterday night, we're lowering DeepSeek V4.1 Abliterated price. $0.12 in $0.025 cache $0.45 output

API issues have been fixed

The discount on DeepSeek V4.1 Abliterated is now permanent. $0.14 in $0.035 cache $0.50 out

More capacity added for DeepSeek

me tonight at the halloween party:

🔥

From AA Cyber benchmark, DeepSeek V4.1 Flash is the third most powerful model for cybersecurity behind only GPT-6 Sol and Kimi K3.1 Pro. DeepSeek beats Grok 4.7, Muse, Mistral and the previous versions of GPT despite it being a fraction of the cost. We serve it abliterated for $0.14 in / $0.50 out / $0.035 cache Zero data retention, no KYC

🔥 Abliterated GLM 5.3 and GLM 5.3 Flash cache rates have been improved by a lot. 98% after 2mins of work

🚀 GLM 5.3 Flash abliterated's cache just dropped. $0.04 → $0.03 If usage keeps rising, we’ll make the lower rate permanent.

DeepSeek V4.1 Flash 1017 OpenRouter: $0.022 / $0.192 / $0.004 Refuseless: $0.01 / $0.08 / $0.002

MiniMax M3.1 Flash Openrouter: $0.06 / $0.24 / $0.015 Refuseless: $0.02 / $0.08 / $0.002

Today's updates > Increased capacity on GLM 5.3 Flash > Fixed thinking traces issue on DeepSeek V4.1

I want no surveillance in my AI just like I want no state looking in my pockets.

Internet should NOT have KYC NO KYC on social media NO KYC on payments (Bitcoin) NO KYC on your data Vive la liberté

We're back online

A short maintenance of the web server is ongoing soon, should last just 5 minutes.

Top 4 uncensored AI models to run locally 😇 > Qwen3.8 27B abliterated > Ornith 1.5 35B A3B abliterated > Cyber-Tiel Coder 35B A3B > MiniCPM5 2B abliterated