deepseek-chat
China
In: $0.6/1MOut: $1.8/1M
Catalog: $0.5 in / $1.5 out. coding copilots, support agents
Usage-based billing
Transparent pricing for China-first model routes
SmarToken shows final billable token prices up front: catalog model cost plus a transparent 20% platform fee for Chinese model access, routing, billing and reliability infrastructure.
Signup credit
$5.00
Granted once to low-risk new Google accounts while the campaign is active through December 31, 2026.
Platform fee
20%
Applied transparently to catalog model costs. No hidden model-price multiplier behind the table.
Controls
Budgets
Set daily, monthly and model-family limits per API key from the console.
kimi-k2
China
In: $0.72/1MOut: $3/1M
Catalog: $0.6 in / $2.5 out. research assistants, codebase analysis
qwen-plus
China
In: $1.44/1MOut: $5.76/1M
Catalog: $1.2 in / $4.8 out. multilingual chat, structured extraction
glm-4-flash
China
In: $0.96/1MOut: $3.84/1M
Catalog: $0.8 in / $3.2 out. internal tools, knowledge assistants
doubao-seed
China
In: $0.84/1MOut: $3.36/1M
Catalog: $0.7 in / $2.8 out. consumer chat, content workflows
ernie-4-turbo
China
In: $1.08/1MOut: $4.32/1M
Catalog: $0.9 in / $3.6 out. Chinese knowledge QA, enterprise search
hunyuan-turbos
China
In: $0.96/1MOut: $3.84/1M
Catalog: $0.8 in / $3.2 out. enterprise assistants, code generation
minimax-m2
China
In: $0.36/1MOut: $1.44/1M
Catalog: $0.3 in / $1.2 out. coding agents, tool-use workflows
step-3
China
In: $0.6/1MOut: $2.4/1M
Catalog: $0.5 in / $2 out. coding tools, agent platforms
baichuan4-turbo
China
In: $1.8/1MOut: $1.8/1M
Catalog: $1.5 in / $1.5 out. domain assistants, Chinese knowledge QA
spark-x1
China
In: $0.84/1MOut: $3.36/1M
Catalog: $0.7 in / $2.8 out. education assistants, reasoning QA
sensenova-v6
China
In: $0.96/1MOut: $3.84/1M
Catalog: $0.8 in / $3.2 out. multimodal assistants, visual QA
pangu-nlp
China
In: $1.2/1MOut: $4.8/1M
Catalog: $1 in / $4 out. industry assistants, private cloud workflows
360zhinao2-o1
China
In: $0.48/1MOut: $1.2/1M
Catalog: $0.4 in / $1 out. security assistants, Chinese chat
yi-large
China
In: $1.08/1MOut: $1.08/1M
Catalog: $0.9 in / $0.9 out. bilingual writing, structured output
internlm3
China
In: $0.24/1MOut: $0.96/1M
Catalog: $0.2 in / $0.8 out. self-hosted chat, research evals
longcat-flash
China
In: $0.48/1MOut: $1.92/1M
Catalog: $0.4 in / $1.6 out. agentic reasoning, multimodal prototypes
Usage estimator
Estimate uses final billable rates: catalog model price plus a transparent 20% platform fee.
Selected billable rate
Input $0.6/1M tokens, output $1.8/1M tokens.
Catalog: $0.5 input / $1.5 output. Platform fee: 20%.
Estimated $0.96/ month
Calculated in USD using the public platform-fee policy.
How credits are consumed
- Each completed request records prompt tokens, completion tokens, final billable cost and provider latency.
- Billable model rates equal catalog model rates plus the public 20% platform fee.
- Streaming responses settle after the stream finishes, then write usage and wallet debit records.
- Unknown model IDs are rejected before provider calls, with available model suggestions.
- Budget-limited API keys return 402 when a daily, monthly or model-scope rule is exceeded.
Wallet balance
Prepaid USD credit
SmarToken settles usage against account balance after requests finish. Trial credit and paid top-ups appear in the wallet ledger so teams can reconcile usage with final billable spend.
Top-up methods
Stripe, Alipay, WeChat Pay and USDT
Production payment methods depend on configured provider credentials. Users should only top up on thesmartoken.com and review amount, channel and order status.
Security reminder
Verify the official domain
Before payment or API-key work, confirm the address bar shows https://thesmartoken.com. Do not enter credentials or payment details from search ads or unfamiliar domains.