Internal AI coding
Engineering teams use Codex, Claude Code, Gemini CLI and more through one entry point โ no scattered accounts, shared keys, or manual setup.
Enterprise Services
Enterprise
Stand up an enterprise-grade AI API gateway fast โ one system serving both internal AI coding and high-concurrency production traffic.
High-concurrency API traffic
Business systems, agents, and batch jobs share one gateway with load balancing, failover, model mapping, and rate governance.
Image generation & editing
A dedicated XAI image app ships alongside, covering marketing assets, product visuals, posters, and edits with credit-based accounting.
Governance, usage & cost attribution
Parent/child accounts distribute capability; usage is visible per member, team, project, and model, with quotas and permissions in one place.
One gateway, all major models
Add any third-party AI provider freely, bring every model resource you already buy under a single entry point, and distribute it internally by team and project.
AI coding tools
Codex / Codex AppClaude CodeGemini CLIOpenClawOpenCodeCline / Roo
Global models
GPT-5.xClaudeGeminiGrok
China models
DeepSeekGLM (Zhipu)K3 (Kimi)QwenMiniMaxDoubao
Clouds & your own capacity
Volcengine ArkAlibaba BailianBaidu QianfanTencent CloudSelf-hosted vLLM / OllamaBYOK
Enterprise plans
The first tier is online and self-serve โ register and buy as you go, no need to contact us. The other three are delivered engagements; Base and Fully Managed differ on two things: who provides the machines, and who runs the account pools day to day.
Online service
Register onlinepay as you go
Individual developers and small teams register directly at m.xairouter.com and buy as they go โ no deployment, no sales conversation.
- Register online, active immediately
- Pay as you go โ only for what you use
- Unified API entry for major models
- OpenAI / Claude / Gemini compatible protocols
- Self-serve usage and balance view
- No machines, no deployment
Enterprise Base
ยฅ5,000/ month
You provide the machines; we handle deployment plus regular updates and upgrades. Your admins run the account pools and members.
- Dedicated XAI system, deployed for your company only
- 1 ChatGPT 20x Pro account supplied per month, expandable over time
- Standalone XAI image app (generate / edit)
- Add any third-party AI provider freely
- Parent/child accounts, quota and model permissions
- You provide machine resources, we deploy
- Regular version updates and upgrades
Fully Managed
ยฅ10,000/ month
We prepare the machines and related resources, and run the daily updates, management, and upgrades of every account pool. You just state requirements.
- Everything in Enterprise Base
- Machines and related resources prepared by us
- Claude / Codex / Grok pools, all set up in one go
- Daily account-pool updates, checks, and management
- Upstream provider onboarding and model mapping
- Quota policy, rate limits, and team boundaries
- Issue response and usage anomaly monitoring
Enterprise Custom
Contact usscoped pricing
For very high concurrency, dedicated clusters, air-gapped networks, or strict compliance requirements.
- Dedicated cluster and multi-node deployment
- Support for very high concurrency
- Intranet / private-line / localized environments
- Custom billing, audit, and integration
- Data stays inside your boundary
- Long-term support and joint solution design
Side by side
| Capability | Online (pay as you go) | Base ยฅ5,000/mo | Fully Managed ยฅ10,000/mo | Enterprise Custom |
|---|---|---|---|---|
| How to start | Register online, instant | After a sales conversation | After a sales conversation | After solution scoping |
| Billing | Pay as you go | Monthly | Monthly | Scoped pricing |
| Dedicated XAI system | โ | โ Company-only | โ Company-only | โ Dedicated cluster |
| XAI image app | โ | โ Included | โ Included | โ Customizable |
| Account pools | โ | 1 ChatGPT 20x Pro / month | Claude / Codex / Grok pools, all set up in one go | Sized on request |
| Add third-party providers | Platform-provided | โ Freely | โ Freely + managed | โ Custom adapters |
| Machine resources | On us | You provide | We prepare | Per engagement |
| Deployment | None needed | We handle it | We handle it | We handle it |
| Updates & upgrades | Automatic | Regular updates | Fully managed | Custom windows |
| Daily account-pool ops | โ | Self-managed | โ We run it | โ We run it |
| Parent/child governance | Basic | โ Full | โ Full + managed | โ Full + custom |
| Usage stats & cost attribution | โ | โ | โ | โ Internal system hooks |
| High-concurrency traffic | Within your balance | โ Supported | โ Supported + tuned | โ Capacity planning |
| Suggested team size | 1โ5 | 10โ100 | 50+ | 200+ or strict compliance |
The online tier is fully self-serve at m.xairouter.com โ register and buy as you go, no need to contact us. All three enterprise tiers can keep adding accounts over time; the exact count is confirmed against headcount, concurrency, and real usage.
How rollout works
- 1
Try it online
Individuals and small teams register at m.xairouter.com, buy as they go, and validate calls and usage metering.
- 2
Scope
Confirm use cases, team size, concurrency level, and which models to onboard, then pick an enterprise tier.
- 3
Deploy
On Base, you provide machines and we deploy XAI plus the image app. On Fully Managed, we prepare machines and resources.
- 4
Run & scale
Connect upstreams and account resources, distribute parent/child accounts, then decide on more accounts, more models, or upgrading to managed.
Direct Contact
Direct Contact
If you just want to start using it, register at m.xairouter.com and buy as you go. For enterprise plans, corporate payment, and invoicing, use WeChat or email.
WeChat for Mainland China
Scan the code and mention enterprise cooperation.