# RunInfra offers flat-rate open models for coding agents

published: 2026-10-20T14:05:51.132Z
topic: infrastructure
source: https://www.producthunt.com/products/runinfra
url: https://aiagentnews.io/stories/runinfra-offers-flat-rate-open-models-for-coding-agents

RunInfra has launched coding plans that let agents such as Claude Code and Codex use open models including GLM 5.3, DeepSeek V4.1 Flash and Qwen3.8 27B for a flat monthly price starting at $10. The company says session-aware routing reaches cache hits up to 99.5% of the time.

RunInfra's coding plans give coding agents access to open models such as GLM 5.3, DeepSeek V4.1 Flash and Qwen3.8 27B for a flat monthly price starting at $10. The plans work with Claude Code, Codex, OpenCode, Cline, Droid and Grok Build. Setup is one prompt pasted into the agent, followed by approving sign-in.

According to the listing, session-aware routing hits cache up to 99.5% of the time, which the company says makes a plan go further. RunInfra also states zero data retention, no training on prompts, and SOC 2 Type II compliance. The product was listed as launched on July 1, 2026.

## Why it matters if you build agents

Developers running coding agents can use these open models under one flat fee instead of paying for tokens as the agent works. Cache hit rates will decide how far a plan stretches across long sessions.

The listing does not say what happens when a user reaches a daily or weekly limit, so check that behavior before moving an agent's workflow over.