Nexum Router FAQ

Nexum Router is a hosted LLM gateway by Dialagram. You subscribe, create an API key in the dashboard, and point your existing tools at one stable endpoint. A flat $5.50/week gets you the same usage as a $200/month Claude or ChatGPT plan, across all 20 models, with no hourly or weekly caps.

Every plan includes $5 of DeepSeek V4.1 Flash usage per day, renewed every day: $35 a week, about $150 a month, at no extra cost. Unused daily usage does not roll over, and you can add more DeepSeek credits from the dashboard.

Nexum Router costs a flat $5.50 per week — about $22 a month. Every model and every supported tool is included in that one price. Billing is weekly with no contract, and you can cancel anytime from the dashboard.

Nexum Router's LLM API pricing is one flat subscription: $5.50 per week, about $22 a month counting four weeks. There is no price per million tokens, no separate input and output rate, and no credit balance to top up for the Qwen, Xiaomi, and Meta models. DeepSeek V4.1 Flash includes $5 of usage a day inside the same price. Create an account at /router/auth, then subscribe and create an API key in the dashboard at /router.

Most LLM APIs bill per token: every request is priced by its input and output tokens, so the bill grows with usage and is hard to predict for coding agents that resend large contexts on every step. Nexum Router charges a flat $5.50/week instead, so a heavy week costs the same as a light one. Pay-per-token billing can cost less for occasional, light use; a flat plan wins for sustained, all-day agentic coding. Plan-by-plan prices and limits for other coding plans are at /alternatives/claude-code.

There is no dollar cap, no hourly limit, no 5-hour session window, and no weekly cap on Qwen, Xiaomi, and Meta models. DeepSeek V4.1 Flash includes $5 of usage a day, renewed daily, with top-ups from the dashboard. What Nexum Router does limit is speed: requests are paced per account so the service stays fast for everyone, and short bursts wait a moment instead of failing. The public status page tracks availability.

20 models across four providers are included: DeepSeek V4.1 Flash (model ID deepseek-v4.1-flash) for fast reasoning and agentic coding, Qwen 3.8 Max (qwen-3.8-max, thinking variant qwen-3.8-max-thinking), which also accepts image input, plus Qwen 3.7 Max (qwen-3.7-max, qwen-3.7-max-thinking) and the wider Qwen 3.5-3.8 family, Xiaomi Mimo 2.6 (xiaomi-mimo-2.6) and Xiaomi Mimo 2.5 (xiaomi-mimo-2.5) for fast iteration, Meta Muse Spark 1.2 (meta-muse-spark-1.2) for end-to-end developer workflows, and Meta Muse Spark 1.3 (meta-muse-spark-1.3) for multimodal reasoning-first coding. Switch models by changing the model ID — nothing else in your setup changes.

Claude Code is supported through a dedicated Anthropic-compatible endpoint: set ANTHROPIC_BASE_URL to the router's /router/claude endpoint with your Nexum API key and keep your exact CLI workflow — slash commands, sessions, everything. Setup takes about two minutes; the step-by-step guide is at /guides/claude-code-api.

Anything that speaks the OpenAI API works: point the tool at /router/v1 as the base URL with your Nexum API key. That covers the official OpenAI SDKs, curl, opencode, OpenClaw, Hermes, Cline, and similar tools. Sign in at /router/auth to create the key; the homepage (/) has copy-paste setup snippets.

Claude Code Max costs $100/month (Max 5x) to $200/month (Max 20x), and both tiers meter usage in 5-hour rolling windows with weekly caps on top. Nexum Router matches or beats the usage ceiling of the top $200/month Max 20x tier for $5.50/week — roughly one-eighth the price — with no windows and no caps, through a native Claude Code endpoint.

Those plans meter usage with dollar caps or monthly credit allowances; Nexum Router is a flat $5.50/week with no usage meter at all. Detailed side-by-side pages: /compare/opencode, /compare/command-code, /compare/cursor, /compare/github-copilot, and the full roundup at /alternatives/claude-code.

In the 7 days ending 2026-08-31, the router served 38.3 billion tokens across 789,056 requests from 83 active users. The median active user routed 43.2 million tokens that week and the average 468 million; the single heaviest user routed 14.3 billion tokens in one week on the same $5.50 plan.

Yes. Billing is weekly with no contract. Cancel from the dashboard whenever you like and you simply are not charged again.

All sales are final. No refunds are granted once you decide to proceed with a purchase. You can cancel anytime from the dashboard to stop future charges.

The public status page at /status shows live provider health for the routed models, updated continuously.

Nexum Router is built by Dialagram, which builds practical AI tooling. Support is at support@dialagram.me, and Dialagram is on X (x.com/Dialagram), LinkedIn, and Facebook.