logo
TokenMoo
HomeConsoleModel SquareRankingsDocsAbout
Sign in
HomeConsoleModel SquareRankingsDocsAbout
Sign in
Independent gateway·Operated by AI CHAMPION LIMITED

Every model. One API. Clear pricing.

Access leading reasoning, coding, image and video models through one OpenAI-compatible endpoint. Compare live rates, including cache pricing, before you spend a token.

Launch pricing: up to 50% off eligible models and periods.

Get Your API KeyRead documentation
New users get $0.5 in free credit
8B+ tokens served daily23,940+ registered users
High cache hit · Rich models

TokenMoo

Claude

Claude

Flagship general-purpose model

BaiLian

Bailian

Video generation model

Qwen

Qwen

Flagship general-purpose model

Z.ai

z-ai

Flagship general-purpose model

OpenAI Compatible
One API
Usage-Based Billing

Start by workflow

What are you building?

Built for the work you do

Code

Build, review, refactor

Office

Docs, analysis, research

Video

Generate scenes & motion

Research

Reason over long context

Live rates · No surprises

Model marketplace

32 models · Prices per 1M tokens · Cache read and write shown when available

⌘K
32models
DeepSeek

deepseek-v4.1-flash

DeepSeek

Efficient multimodal model designed for fast inference, high throughput, agentic tasks, and cost-effective AI applications.

Dynamic Pricing50% OFF
Input$0.3→$0.15/ 1M
Output$1.2→$0.6/ 1M
Cache Read$0.006→$0.003/ 1M
Current period price
ProtocolsOpenAI
Z.ai

glm-5.3

z-ai

Flagship coding and agentic model built for complex software engineering, long-horizon tasks, and advanced reasoning.

Dynamic Pricing40% OFF
Input$1.4→$0.84/ 1M
Output$4.4→$2.64/ 1M
Cache Read$0.28→$0.168/ 1M
ProtocolsOpenAI
MoonshotAI

kimi-k3

Moonshot

Frontier multimodal model with a 1M-token context window, built for long-horizon coding, complex reasoning, and knowledge work.

Dynamic Pricing40% OFF
Input$3→$1.8/ 1M
Output$15→$9/ 1M
Cache Read$0.3→$0.18/ 1M
ProtocolsOpenAI
Z.ai

glm-5.3-flash

z-ai

Efficient multimodal model combining strong coding and agentic capabilities with low-cost inference and long-context understanding.

Dynamic Pricing40% OFF
Input$0.15→$0.09/ 1M
Output$0.5→$0.3/ 1M
Cache Read$0.03→$0.018/ 1M
ProtocolsOpenAI
Claude

claude-fable-5

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$10→$5/ 1M
Output$50→$25/ 1M
Cache Read$1→$0.5/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-fable-5-1

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$10→$5/ 1M
Output$50→$25/ 1M
Cache Read$0.25→$0.125/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-opus-4-5

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$5→$2.5/ 1M
Output$25→$12.5/ 1M
Cache Read$0.5→$0.25/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-opus-4-6

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$5→$2.5/ 1M
Output$25→$12.5/ 1M
Cache Read$0.5→$0.25/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-opus-4-7

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$5→$2.5/ 1M
Output$25→$12.5/ 1M
Cache Read$0.5→$0.25/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-opus-4-8

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$5→$2.5/ 1M
Output$25→$12.5/ 1M
Cache Read$0.5→$0.25/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-opus-5

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$5→$2.5/ 1M
Output$25→$12.5/ 1M
Cache Read$0.5→$0.25/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-opus-5-5

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$4→$2/ 1M
Output$20→$10/ 1M
Cache Read$0.2→$0.1/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-sonnet-4-6

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$3→$1.5/ 1M
Output$15→$7.5/ 1M
Cache Read$0.3→$0.15/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-sonnet-5

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$2→$1/ 1M
Output$10→$5/ 1M
Cache Read$0.2→$0.1/ 1M
ProtocolsAnthropicOpenAI
Claude

claude-sonnet-5-5

Claude

Flagship general-purpose model · Code generation, Agents & tool use

Dynamic Pricing50% OFF
Input$2→$1/ 1M
Output$10→$5/ 1M
Cache Read$0.1→$0.05/ 1M
ProtocolsAnthropicOpenAI
DeepSeek

deepseek-v4-pro

DeepSeek

Flagship reasoning model for complex agentic workflows, advanced coding, mathematical reasoning, and knowledge-intensive tasks.

Dynamic Pricing30% OFF
Input$1.32→$0.924/ 1M
Output$3.96→$2.772/ 1M
Cache Read$0.132→$0.0924/ 1M
Current period price
ProtocolsOpenAI
Z.ai

glm-5.1

z-ai

A general-purpose language model designed for conversational AI, code generation, reasoning, and everyday AI applications.

Dynamic Pricing30% OFF
Input$1.4→$0.98/ 1M
Output$4.4→$3.08/ 1M
Cache Read$0.26→$0.182/ 1M
ProtocolsOpenAI
Z.ai

glm-5.2

z-ai

A powerful language model designed for advanced reasoning, complex coding tasks, and agentic workflows, with strong capabilities in software engineering and problem-solving.

Dynamic Pricing40% OFF
Input$1.4→$0.84/ 1M
Output$4.4→$2.64/ 1M
Cache Read$0.28→$0.168/ 1M
ProtocolsOpenAI
MoonshotAI

kimi-k2.6

Moonshot

Agentic AI model designed for coding, visual understanding, and long-running tasks, with a focus on fast everyday interactions.

Dynamic Pricing40% OFF
Input$0.95→$0.57/ 1M
Output$4→$2.4/ 1M
Cache Read$0.16→$0.096/ 1M
ProtocolsOpenAI
MoonshotAI

kimi-k2.7-code

Moonshot

A coding-focused AI model designed for software development, code generation, debugging, and agentic programming workflows.

Dynamic Pricing30% OFF
Input$0.95→$0.665/ 1M
Output$4→$2.8/ 1M
Cache Read$0.19→$0.133/ 1M
ProtocolsOpenAI

Page 1 of 2

Service provider and brand relationship

TokenMoo is operated by AI CHAMPION LIMITED

TokenMoo is an independent AI API gateway. It provides API access to supported model services including DeepSeek, Kimi, GLM, Qwen and more, and is not the official website or direct retail service of any model provider. Model names and trademarks belong to their respective owners.

Supported model services: DeepSeek, Kimi, GLM, Qwen and more

Who you buy from

TokenMoo provides the account, billing and support service through AI CHAMPION LIMITED.

What you buy

Usage-based API access to the supported model catalogue through one compatible endpoint.

Brand scope

Provider names describe compatible or supported model services and do not imply official provider affiliation.

Pricing conditions

Rates may vary by model, account group and request time. Review the current model pricing before purchasing.

About TokenMooUser AgreementPrivacy PolicyRefund PolicyContact support

Questions, answered

Why route your model traffic through this gateway, how billing works, and what to expect.

Why buy through this gateway instead of each provider directly?

Two practical reasons. Aggregation: one API key and one OpenAI-compatible endpoint give you GLM, Kimi, DeepSeek and other models, so you stop juggling separate accounts, keys and invoices. Payment: you can pay with the local methods shown on the top-up page instead of arranging an overseas card for every provider. Pricing stays transparent — every model price is published on the model square, so you can compare before you spend.

How is pricing calculated?

Usage is billed as you go — per token, with the unit price listed on each model page. You top up a balance and usage is deducted from it.

Which payment methods are supported?

The top-up page lists the channels currently enabled for this site, including local QR-code payments where available. The methods shown there are always the authoritative list.

Can one API key use every model?

Yes. Point any OpenAI-compatible client at this gateway and use the same key. Which models a key can reach depends on the groups your account is allowed to use and the balance available.

Which clients and SDKs work with it?

Anything that speaks the OpenAI or Anthropic protocol works, for example Claude Code, Cline, OpenCode, Kilo Code and Cherry Studio, plus the official SDKs of those ecosystems.

How do I request a refund?

Refund eligibility and the payout window are described in the refund policy; the same page explains how to reach support with your order details.

How is my data handled?

Request handling and account data follow the privacy policy published on this site.

Pricing notices

  • ·Compatible with OpenAI-, Anthropic- and Gemini-style APIs, so existing SDKs, CLI tools and gateways keep working after a single base URL change.

Refund PolicyPrivacy PolicyUser Agreement

claude-opus-5

Model details

Claude

claude-opus-5 API pricing and compatibility

Claude·Dynamic Pricing

Flagship general-purpose model · Code generation, Agents & tool use

TPS
—
Average latency
—
Success rate
—

Pricing

Base Price

Input
$5/ 1M
Output
$25/ 1M
Cache Read$0.5/ 1M
Cache Write (5m)$6.25/ 1M
Cache Write (1h)$10/ 1M
Dynamic Pricing
Prices vary by usage tier and request conditions
Tiered price table
standard
Input
$5.0000
Output
$25.0000
Cache Read
$0.5000
Cache Write
$6.2500
Cache Write (1h)
$10.0000
TierInputOutputCache ReadCache WriteCache Write (1h)
standard
$5.0000
$25.0000
$0.5000
$6.2500
$10.0000

Pricing by Group

0.5x0.5x
TierInputOutputCache ReadCache Write (5m)Cache Write (1h)
standard
$2.5
$12.5
$0.25
$3.125
$5

Prices shown per 1M tokens

Model

ProviderClaude
TypeDynamic Pricing
Groups
0.5x
Endpoints
anthropicopenai

Copy model name