API platform serving fast open-weight models with OpenAI/Anthropic endpoints for coding agents.
RunInfra homepage
1 / 2Overview
What is RunInfra?
Fast open models for your coding agent.
RunInfra is an inference platform that serves fast open-weight models (GLM, DeepSeek, Qwen and others) through OpenAI- and Anthropic-compatible APIs, built for AI coding agents and developers.
One API key works with both the OpenAI and Anthropic SDKs. Session-aware routing keeps repeat calls close to cached context, reaching around 99% cache hits so cached tokens are billed at lower cached prices. Coding Plans connect 19 coding agents such as Claude Code, Codex, OpenCode, Cline and Droid to open models at a fixed monthly price, with usage resetting every 5 hours and weekly. The model library publishes open per-token prices, cache hit rates and speed metrics.
Billing is flexible: pay-as-you-go credits (minimum $5 top-up, valid one year) or Coding Plans from $10/month, plus custom enterprise tiers. New accounts get $1 in free credit that never expires. By default RunInfra keeps zero data retention, encrypts data in transit and at rest, does not train on user data, and holds SOC 2 Type II attestation.
Platforms and languages
RunInfra Availability
Platforms
Languages
Capabilities
RunInfra Key Features
OpenAI/Anthropic-compatible hosted model APIs
One API key works with both the OpenAI and Anthropic SDKs, so existing code and agents connect without rewrites.
Session-aware cache routing
Session-aware routing sends repeat calls to the replica holding cached context, reaching about 99% cache hits billed at lower cached prices.
Coding Plans for AI coding agents
Fixed monthly plans connect 19 coding agents (Claude Code, Codex, OpenCode, Cline, Droid and more) to open models, with usage resetting every 5 hours and weekly.
Open model library with transparent metrics
The model library shows open per-token prices, cache hit rates and speed metrics for models such as GLM 5.3, DeepSeek V4.1 Flash and Qwen3.8.
Zero data retention and SOC 2 Type II
Zero data retention by default, TLS in transit, AES-256 at rest, no training on user data, and SOC 2 Type II attestation.
Best for
Who uses RunInfra?
Developers using AI coding agents such as Claude Code, Codex or OpenCode
Plug agents into fast, low-cost open models without managing inference infrastructure.
Teams building apps on open-weight LLMs
Get one API key with transparent per-token pricing, high cache hit rates and zero data retention.
Plans and access
RunInfra Pricing
Freemium
Pay-as-you-go from $5 top-up; Coding Plans $10/$29/$99 per month; $1 free credit on signup.
Common questions
RunInfra FAQs
What does a Coding Plan cover?
Coding Plans cover chat completions, messages and responses calls within your workspace. Any key not marked Credits only can use the plan. Image and audio calls are always billed pay-as-you-go.
What happens when I reach my plan limit?
If you enable "Keep full speed on credits", the plan continues on credits at your chosen cap. Once credits run out, a slower free Standby takes over. If neither is available, requests stop and explain why, when limits reset and what to do next.
Can I cancel my plan?
You can cancel anytime from Billing; the plan runs until the end of the current period. Deleting your account ends the plan immediately with no refund for the current month, but your keys keep working on any remaining credit balance.
How does billing work?
Signing up gives you $1 in free credit that never expires. Top-ups start at $5, one credit equals one US dollar, and purchased credits are valid for one year.
Reviews
See what the community thinks and share your experience.
—
Based on 0 ratings
Rating distribution
Community reviews
No reviews
No written reviews yet.
