OpenDesign CoRouter | Unified AI Hub for Teams

One credit pool. A whole team. Lower costs, clear budgets.

Share model credits, set a budget for each member and manage spending together. See the difference in one team example.

Before · 5 plans$500/ month

After · 1 shared plan$100/ month

80% lower cost

Save $400 every month

One subscription for the whole team, unlimited seats

Allocate by need. See individual usage.

5 members

Shared credits / month$120

Used this month$74

Remaining$46

Member and usageAllocated budget

Lin · Engineering lead$40
$32 used$8 left
Wang · Developer$30
$18 used$12 left
Chen · Designer$24
$9 used$15 left
Zhou · Product manager$16
$12 used$4 left
Shen · Operations$10
$3 used$7 left

Member budgets add up to the shared pool. Each key tracks usage separately.

Your models. Your agents. One place.

One account connects model access, team credits and the tools you already use.

Models & credits

CoRouter plans and credits

30+ supported modelsModels for coding, reasoning and multimodal tasks, with access based on your plan.

Shared plan-model creditsOne team pool with per-member budgets. Coming soon.

Personal & team Agent subscriptionsComing soonManage your existing personal and team Agent subscriptions in one place.

Model accessPermissions follow your plan
Member keysIndependent access for each member
Credit allocationShared pool, clear member budgets
Usage & spendingManage your AI work in one place

Work in your familiar tools

Coding, design, collaboration and applications.

Coding AgentCodex · Claude Code · OpenCode · Hermes

IM & teamworkDiscord · Slack · Feishu

Office & designDocuments · Visual concepts · Prototypes

API & applicationsResponses · Messages · Chat Completions

Model families

GPT
DeepSeek
MiMo
Hunyuan
Claude
Kimi
Gemini
Grok
GLM
Qwen

Supported agents

Aider
Antigravity
Claude Code
Codex
GitHub Copilot CLI
Cursor
DeepSeek Harness
Devin for Terminal
Gemini CLI
Grok Build
Hermes
Kilo
Kimi CLI
Kiro CLI
OpenCode
Pi
Qoder CLI
Qwen Code
DeepSeek Reasonix
Trae CLI
Mistral Vibe CLI

See how each model gets used.

Explore real token shares by day or week, and see how model usage changes.

Model usage share

Select a bar to see each model’s share of tokens that day or week.

Last 8 weeksAug 17 – Oct 11 · UTC

DeepSeek V4.1 FlashDeepSeek V4 FlashDeepSeek V4 ProDeepSeek V4 Flash Vision ExpGLM-5.3 FlashGLM-5.2MiMo V2.6 ProGPT-6 LunaOther models

Model rankingsShare of period

Your models. Your workflow.

Access GPT, Claude, DeepSeek, Gemini, Grok, Kimi, Qwen, GLM, MiMo and Hunyuan through one entry point.

GPT-6.1 Sol

New

Up to 272,000 tokens

Input / 1M tokens: $2
Output / 1M tokens: $10
Cache read / 1M tokens: $0.1
Cache write / 1M tokens: $2.5

Over 272,000 tokens

Input / 1M tokens: $4
Output / 1M tokens: $15
Cache read / 1M tokens: $0.2
Cache write / 1M tokens: $5

See details

All 33 models

Get connected in three steps.

01

Get your API key

Choose a plan and create your API key in your workspace.

02

Connect through CC Switch / Magpie

Choose your agent in Quick setup and import to CC Switch, or add OpenDesign as a provider in Magpie.

03

Start using your agent

Activate the configuration and start a new agent session to use your model.

One gateway. Wherever you work.

Connect the tools you use every day to your team’s models.

Use your existing coding agents.

Connect Claude Code, Codex, OpenCode and other compatible tools. Each tool keeps its own key and model configuration.

Choose a setup method to connect your familiar agents.

Installation & setup guide

Get connected

Install. Import. Start working.

01
Install CC Switch, or skip if you already have it.
macOS · requires Homebrew
```
brew install --cask cc-switch
```
Get an API key, choose your agent in Quick setup, then import into CC Switch and enable it.
Start a new session in your agent.

Claude Code

Codex

OpenCode

Hermes

Gemini CLI

Grok Build

Pi

Set up once. Connect your agents.

01
Install Magpie, or skip if you already have it.
macOS / Linux · terminal
```
curl -fsSL https://usemagpie.ai/install.sh | sh
```
Add OpenDesign as a provider with your API key and endpoint.
Pick a model for your agent and start a new session.

Claude Code

Codex

OpenCode

Hermes

Gemini CLI

Grok Build

Pi

Cursor

Kimi CLI

DeepSeek Harness

Connect models to your products and workflows.

Use an API key and compatible endpoints to connect SDKs, custom apps, automations and other agents. Choose Chat Completions, Responses or Messages according to the model’s supported protocols.

API KeySDKAutomationCustom apps

Choose a connection method in your workspace. Each resource keeps its own billing and access scope.

Create an app API key

your-app.ts

const client = new OpenAI({
  baseURL: COROUTER_BASE_URL,
  apiKey: process.env.COROUTER_API_KEY,
});

// Choose a model available to your team
await client.chat.completions.create({
  model: YOUR_MODEL_ID,
  messages,
});

Get started from $8.
Up to 15× the credit value.

Choose a plan for your usage. See model access and included credits at a glance.

First-month offer

20% off 1st month30% off

Light needs · Easy delivery

$8/ mo~~$10~~First-month price

$7/ mo~~$10~~

Save $36Billed $84/year

Subscribe now

Total monthly credits, up to$120

Design Plan Up to $120 per month

Other benefits

Design Plan

Up to $120 per month

Claude Haiku 5.5
GPT-6 Luna
DeepSeek V4.1 Flash
GLM-5.3 Flash-X
MiMo V2.6 Flash
MiMo V2.6 Pro
Hunyuan H4 Preview
DeepSeek V4 Flash
DeepSeek V4 Flash Vision Exp
DeepSeek V4 Pro
GLM-5.3 Flash
Kimi K2.7 Code

Plan models

Not included

GPT-6 Astra
Claude Fable 5.1
Claude Opus 5.5
GPT-6 Sol
Muse Spark 1.3
Gemini 4 Argon
Kimi K3
Grok 4.7
Gemini 3.8 Flash
Claude Sonnet 5.5
Seedream 5 Lite / Pro
Nano Banana 2
GPT Image 2

✓OpenDesign API access
✓10 concurrent tasks
✓Zero-config professional design agent
✓Publish artifacts online and share them
✓162+ Skills workflows
✓152+ Design Systems
✓Email support

20% off 1st month30% off

Everyday design · Continuous delivery

$16/ mo~~$20~~First-month price

$14/ mo~~$20~~

Save $72Billed $168/year

Subscribe now

Total monthly credits, up to$180

Design Plan up to $160+ plan models $20

Other benefits

Design Plan

Up to $160 per month

Claude Haiku 5.5
GPT-6 Luna
DeepSeek V4.1 Flash
GLM-5.3 Flash-X
MiMo V2.6 Flash
MiMo V2.6 Pro
Hunyuan H4 Preview
DeepSeek V4 Flash
DeepSeek V4 Flash Vision Exp
DeepSeek V4 Pro
GLM-5.3 Flash
Kimi K2.7 Code

Plan models

Standard · $20/month

GPT-6 AstraNot included
Claude Fable 5.1Not included
Claude Opus 5.5Not included
GPT-6 Sol
Muse Spark 1.3
Gemini 4 Argon
Kimi K3
Grok 4.7
Gemini 3.8 Flash
Claude Sonnet 5.5
Seedream 5 Lite / ProUp to 1K
Nano Banana 2Up to 1K
GPT Image 2Up to 1K

✓OpenDesign API access
✓20 concurrent tasks
✓Zero-config professional design agent
✓Publish artifacts online and share them
✓162+ Skills workflows
✓152+ Design Systems
✓Email support

Design Plan

Up to $360 per month

Claude Haiku 5.5
GPT-6 Luna
DeepSeek V4.1 Flash
GLM-5.3 Flash-X
MiMo V2.6 Flash
MiMo V2.6 Pro
Hunyuan H4 Preview
DeepSeek V4 Flash
DeepSeek V4 Flash Vision Exp
DeepSeek V4 Pro
GLM-5.3 Flash
Kimi K2.7 Code

Plan models

6× · $120/month

GPT-6 Astra
Claude Fable 5.1
Claude Opus 5.5
GPT-6 Sol
Muse Spark 1.3
Gemini 4 Argon
Kimi K3
Grok 4.7
Gemini 3.8 Flash
Claude Sonnet 5.5
Seedream 5 Lite / ProUp to 2K
Nano Banana 2Up to 2K
GPT Image 2Up to 2K

✓OpenDesign API access
✓30 concurrent tasks
✓Zero-config professional design agent
✓Publish artifacts online and share them
✓162+ Skills workflows
✓152+ Design Systems
✓Priority email support

Lowest unit price

40% off 1st month51% off

High-volume creation · Consistent output

$120/ mo~~$200~~First-month price

$98/ mo~~$200~~

Save $1,224Billed $1,176/year

Subscribe now

Total monthly credits, up to$900

Design Plan up to $600+ plan models $300

Other benefits

Design Plan

Up to $600 per month

Claude Haiku 5.5
GPT-6 Luna
DeepSeek V4.1 Flash
GLM-5.3 Flash-X
MiMo V2.6 Flash
MiMo V2.6 Pro
Hunyuan H4 Preview
DeepSeek V4 Flash
DeepSeek V4 Flash Vision Exp
DeepSeek V4 Pro
GLM-5.3 Flash
Kimi K2.7 Code

Plan models

15× · $300/month

GPT-6 Astra
Claude Fable 5.1
Claude Opus 5.5
GPT-6 Sol
Muse Spark 1.3
Gemini 4 Argon
Kimi K3
Grok 4.7
Gemini 3.8 Flash
Claude Sonnet 5.5
Seedream 5 Lite / ProUp to 4K
Nano Banana 2Up to 4K
GPT Image 2Up to 2K

✓OpenDesign API access
✓30 concurrent tasks
✓Zero-config professional design agent
✓Publish artifacts online and share them
✓162+ Skills workflows
✓152+ Design Systems
✓Peak-time priority compute · lower latency
✓Dedicated customer success
✓Peak-time priority compute · lower latency
✓Creator / developer support

Reliability, control, data and billing.

99.99% SLA. Automatic failover.

Route to a backup provider when an upstream service fails. Availability and remedies follow the applicable service agreement.

Members, keys and budgets. One control plane.

Assign independent keys and spending limits, review usage by member, and revoke access when someone leaves.

Usage and billing records. No retained content.

Record the member, model, token usage and cost of each request. Prompt and response content is not stored or used to train models.

Upstream failure? No charge.

Credits are reserved when a request starts and settled after a complete response. Upstream errors release the reserved amount.

Know where your models come from.

See the provider, exact model ID and pricing in the model catalog. Access models included in your plan.

Compared with hosted routers, self-hosted gateways and resellers.

Compare cost, control and setup at a glance.

Get API key

Dimension	CoRouter	OpenRouter & others	Self-hosted	Resellers / sharing
Cost	Go: $8 first month, up to $120 credits	Model rates + 5.5% fee¹	Model rates + operations	Operator-specific pricing
Billing	Member usage and billing together	Prepaid, usage-based billing	Consolidate provider bills yourself	Operator-specific policies
Reliability	99.99% SLA · automatic failover	Provider fallback; enterprise SLA	Manage monitoring and failover	Verify SLA terms
Model supply	Managed supply; provider rules apply	Published providers and policies	Your own provider accounts	Verify supply and resale rights
Team budgets	Individual keys · shared budgets	Workspace and key limits	Configure permissions yourself	Verify member controls
Data privacy	Platform and provider privacy terms	Policy routing; zero-retention options	Your storage and logging	Verify retention and training policies
Agent setup	20+ agents · centralized setup	Compatible APIs and setup guides	Deploy, then configure agents	Manual endpoint and key setup
Your resources	Model credits in one place	BYOK supported	Bring provider keys	BYOK varies by operator
Deployment	Self-serve · enterprise inquiries	Managed service · enterprise procurement	Deploy and maintain yourself	Verify contracts and invoices

Public pricing references · 11 Oct 2026: OpenRouter Vercel Requesty LiteLLM Model offer

¹ OpenRouter Standard fee. Go offer: first month and eligible models only. Team features, SLA and provider policies follow service terms; external subscriptions remain separate.

Real work. Shipped with AI.

See how different teams organize model subscriptions, shared budgets and everyday AI workflows.

One subscription, cost control per task

Email marketing, smart copywriting and customer intent analysis for B2B brands. The team bought OpenAI, Anthropic and Mistral separately, maintained several sets of keys and bills, and switched models by hand at peak times.

European SaaS marketing automation companyBerlin, Germany | 128-person team

One budget, less time reconciling

Product descriptions, multilingual ads, review analysis and image prompts generated in bulk for multiple DTC stores. The team used models from five vendors, with budgets scattered and token usage hard to see in one place.

North American cross-border e-commerce content groupToronto, Canada | 276-person team

Automatic model fallback, always-on service

AI customer-service knowledge bases and document summaries for small businesses in Southeast Asia. A limited budget couldn’t cover several commercial subscriptions, and rate limits on a single model hit the live product and customer experience directly.

Southeast Asian enterprise AI startupSingapore | 46-person team

Before your first request.

Do I need a separate model plan for every teammate?

Once team sharing launches, members will be able to use a shared model-credit pool through individual API keys and member budgets. Seat and credit rules will be confirmed at launch. Agent software subscriptions are billed separately.

How do shared credits and model offers work?

Team members share the plan’s model-credit pool. Model-specific promotions do not become unrestricted credits for every model. Check the selected plan’s access rules before connecting.

Do invited members each get another credit grant?

No. Membership is free; credit grants follow the purchased plan. Inviting members does not multiply the plan’s allowance.

Can I pool Claude or ChatGPT subscriptions?

No. Third-party personal subscriptions remain for their account holder. The shared pool here refers to your CoRouter plan-model credits.

How do monthly and yearly prices work?

Monthly cards show first-month offers of $8, $16, $70 and $120 for Go, Plus, Pro and Max. Renewals are $10, $20, $100 and $200 per month. Annual plans charge $84, $168, $720 or $1,176 upfront; the monthly equivalent is not a monthly charge.

Do connected agents and API calls use the same credits?

Eligible agent and API usage draws from the same CoRouter plan. Model access, credit pools and usage windows follow your subscription. Free usage and limited-time in-app offers do not automatically apply to API, MCP or CLI usage.

Save 80% on costs

One subscription, unlimited seats for your whole team.

One subscription for the whole team, unlimited seats

Allocate by need. See individual usage.

Models & credits

Work in your familiar tools

See how each model gets used.

Model usage share

Your models. Your workflow.

Up to 272,000 tokens

Over 272,000 tokens

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

Up to 272,000 tokens

Over 272,000 tokens

No input-length tiers

No input-length tiers

No input-length tiers

Up to 272,000 tokens

Over 272,000 tokens

Up to 100,000 tokens

Over 100,000 tokens

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

Up to 272,000 tokens

Over 272,000 tokens

No input-length tiers

No input-length tiers

No input-length tiers

Up to 272,000 tokens

Over 272,000 tokens

Up to 272,000 tokens

Over 272,000 tokens

Up to 272,000 tokens

Over 272,000 tokens

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

No input-length tiers

Get connected in three steps.

Get your API key

Connect through CC Switch / Magpie

Start using your agent

One gateway. Wherever you work.

Use your existing coding agents.

Install. Import. Start working.

Set up once. Connect your agents.

Delegate tasks to an agent in your team chat.

Create and deliver inside OpenDesign.

Connect models to your products and workflows.

Get started from $8.Up to 15× the credit value.

Reliability, control, data and billing.

99.99% SLA. Automatic failover.

Members, keys and budgets. One control plane.

Usage and billing records. No retained content.

Upstream failure? No charge.

Know where your models come from.

Real work. Shipped with AI.

One subscription, cost control per task

One budget, less time reconciling

Automatic model fallback, always-on service

Get your API key.Bring your team’s AI together.

Before your first request.

Get started from $8.
Up to 15× the credit value.

Get your API key.
Bring your team’s AI together.