Just tried the latest deepseek_ai V4-013 model using the DeepSeek harness through nebiustf TF Relay.
The entire setup and integration experience was seamless, and the model feels impressively quick in real-world use. It handled the task really well while maintaining a strong
Another Astra + Deepseek 4.1 Flash test!
Pretty cool game 100% autonomously generated by - 0 guidance. Still a long way to go - 2h time, 87% of the cost being Astra overhead that I'm sure can be reduced. Next goal: 30min max.
I finally managed to build my own tiny GTA with DeepSeek V4.1 Flash 😭
Steal a car. grab the package. lose the cops. traffic, handbrake turns, wanted stars. all playable in the browser. “just one more test drive” is becoming a problem.
You can now run Hermes Agent with DeepSeek V4.1 Flash through Token Harbour and give your agent a…
You can now run Hermes Agent with DeepSeek V4.1 Flash through Token Harbour and give your agent a powerful AI model without paying for expensive API usage. With this setup, you can: 🤖 Run Hermes Agent with DeepSeek 🧠 Give Hermes a powerful AI “brain” 💸 Use a free API option…
You have reached the end of the archive
All of deepseek