Pay per token. No subscription required. Optional Team and BYOK plans.
10 low-cost trial calls on signup · +100 bonus on first $5 top-up · No card required
Showing 16 of 16 models
Nearly Free
$0.01/M
Budget
$0.07–3.00/M
Mid
$0.24–4.00/M
Premium
$1.50–30.00/M
| Model | Provider | Input $/M | Output $/M | Context | Tier |
|---|---|---|---|---|---|
| GLM-4 Flashzhipu/glm-4-flash | Zhipu | $0.01 | $0.01 | 128K | Nearly Free |
| Doubao Pro 256Kdoubao/doubao-pro-256k | ByteDance | $0.07 | $0.12 | 256K | Budget |
| Qwen Turboqwen/qwen-turbo | Alibaba | $0.08 | $0.31 | 128K | Budget |
| Gemini 2.5 Flashgoogle/gemini-2.5-flash | $0.36 | $3.00 | 1M | Budget | |
| Qwen Plusqwen/qwen-plus | Alibaba | $0.13 | $1.87 | 128K | Budget |
| Moonshot v1 8Kmoonshot/moonshot-v1-8k | Moonshot | $0.14 | $0.14 | 8K | Budget |
| GPT-4o Miniopenai/gpt-4o-mini | OpenAI | $0.18 | $0.72 | 128K | Budget |
| Kimi K2.5moonshot/kimi-k2.5 | Moonshot | $0.24 | $1.20 | 256K | Mid |
| DeepSeek V3deepseek/deepseek-chat | DeepSeek | $0.34 | $0.50 | 64K | Mid |
| DeepSeek R1deepseek/deepseek-reasoner | DeepSeek | $0.34 | $0.50 | 64K | Mid |
| MiniMax Text 01minimax/minimax-text-01 | MiniMax | $0.36 | $1.44 | 1M | Mid |
| GLM-5.1zhipu/glm-5.1 | Zhipu | $1.25 | $4.00 | 128K | Mid |
| Gemini 2.5 Progoogle/gemini-2.5-pro | $1.50 | $12.00 | 1M | Premium | |
| GPT-5openai/gpt-5 | OpenAI | $3.00 | $18.00 | 272K | Premium |
| Claude Sonnet 4anthropic/claude-sonnet | Anthropic | $3.60 | $18.00 | 200K | Premium |
| Claude Opus 4.6anthropic/claude-opus | Anthropic | $6.00 | $30.00 | 200K | Premium |
Foreign model prices include gateway access and failover · Chinese model prices ≈ direct provider + small aggregation fee · All billed via AI POWER LIMITED (HK).·$5 first top-up: zero fee, exact $5 charged·+5.5% Stripe fee on $20+ tiers
Price alone doesn't win you over. These do:
Qwen + DeepSeek + GLM + Kimi alongside GPT + Claude + Gemini. No competitor has this in 2026.
China merchants pay in CNY. Enterprise gets fapiao (专票/普票). No foreign card required.
OpenAI 5xx? Request transparently routes to Claude/DeepSeek. model="auto-cheap" saves 60-80%.
No card required · sign up with email or Google
1
Add $5+ via credit card, WeChat Pay, or Alipay. First $5+ top-up gets +100 bonus calls. Credits never expire.
2
Each API call deducts based on actual token usage. No rounding up. Smart routing lets you save 60-80%.
3
No subscription required for self-serve credits. Custom amounts start at $5. Add credits when you need them.
Use model="auto-cheap" to automatically route to the cheapest suitable model. Typical cost for 1M mixed requests:
| Strategy | Cost / 1M requests |
|---|---|
| Claude Opus for everything | $17,250 |
| GPT-5 for everything | $10,065 |
| AIPower smart routing (auto-cheap 80% · auto-code 15% · auto-best 5%) | $1,200 |
| Savings vs Opus | 93% |
10 low-cost trial calls. +100 bonus on first $5 top-up. From $0.01/M tokens.