Providers

OpenUsage supports 19 providers spanning local coding agents and cloud API platforms. Most are auto-detected on first run; the rest need a single environment variable. Each tile on the dashboard maps to one provider page below.

Coding agents

These providers read local files, OAuth credentials, or shell out to a CLI. No API key is required for most of them.

Claude CodeSessions, billing blocks, burn rate, per-model tokensCursor IDEPlan spend, billing cycle, composer sessions, AI code scoreGitHub CopilotChat/code/premium quotas, org seats, rate limitsCodex CLISessions, rate-limit windows, credit balance, planGemini CLIOAuth status, session tokens, MCP config, user quotaOpenCodeZen models, spend via telemetry plugin

Local runtimes

Self-hosted model servers running on this machine.

OllamaLocal models, VRAM, request log analytics, cloud credits

API platforms

These providers require an API key in an environment variable. Some return only rate-limit headers, others return full billing and usage data.

OpenAIRPM/TPM rate limitsAnthropicRPM/TPM rate limitsOpenRouterCredits, daily/weekly/monthly usage, generation analytics, BYOKGroqRPM/TPM/RPD/TPD rate limitsMistral AIMonthly budget, credit balance, spend, tokens (EUR)DeepSeekBalance breakdown, rate limits (CNY)MoonshotBalance breakdown, quotas, peak usage (USD or CNY)PerplexityPro / Max plan quotas via browser-session authxAI (Grok)Credits, rate limits, allowed models (USD)Z.AI5h window, monthly usage, credit grants, tool usageGemini APIModel catalog, per-model token limitsAlibaba Cloud Model StudiosBilling period, balance, spend, per-model quotas (USD)