River AI gives new users $25 in free API credits for DeepSeek V4.1 Flash and GLM-5.3 Flash

River AI builds a training and inference platform for open-weight models. On October 11 it launched River Inference, and everyone who signs up gets $25 in free credits. You can spend them on DeepSeek V4.1 Flash, GLM-5.3 Flash, and other models through an OpenAI-compatible API, so it drops straight into the coding tools and chat clients you already use.

What you get

• New River accounts come with $25 in free credits for the inference API.
• The current lineup is DeepSeek V4.1 Flash (plus a separate 1017 release), GLM-5.3 Flash, and NVIDIA Nemotron 3.5 Super VL. All four have a context window of about 1M tokens and support image input, streaming, reasoning, and function calling.
• DeepSeek V4.1 Flash costs $0.12 per 1M input tokens and $0.48 per 1M output tokens, and GLM-5.3 Flash costs $0.12 input and $0.40 output. At those rates, $25 covers roughly 200M input tokens, or around 50M DeepSeek output tokens.

How to use

1. Sign up at the River Console and create an API key on the API Keys page.
2. In any OpenAI-compatible client or coding tool, set the base URL to https://api.river.ai/v1 and pick deepseek-ai/DeepSeek-V4.1-Flash-1017 or zai-org/GLM-5.3-Flash as the model.

See the announcement in Igor Babuschkin's post on X, models and prices in the River models and pricing docs, and setup details in the River inference docs.