China-AI-Pricing-FreeTier-Tracker

What does China AI Pricing & Free Tier Tracker do?

China AI Pricing & Free Tier Tracker is an Apify Actor that continuously monitors 11 major Chinese AI providers and delivers structured, machine-readable data on API pricing, free tiers, promotions, and price changes.

Instead of manually checking dozens of Chinese AI pricing pages, you get structured datasets updated on your schedule — ready for cost optimization, model routing, AI FinOps, or building your own pricing dashboard.

Providers covered: DeepSeek, Alibaba/Qwen, Zhipu AI/GLM, Moonshot/Kimi, Baidu/ERNIE, MiniMax, ByteDance/Doubao, Tencent/Hunyuan, Baichuan, 01.AI/Yi, SiliconFlow

Key feature: This Actor tracks 4 types of free models — permanently free, limited-time free, public beta, and trial credits — with confidence levels and change detection.

Why use this Actor?

• Save hours of manual research — Chinese AI pricing is fragmented across 10+ platforms, all in Chinese
• Track free tiers — find which models offer free quotas and how much
• Never miss a deal — promotions, discounts, and limited-time offers are captured automatically
• Detect price changes — scheduled runs compare against historical snapshots and flag changes
• Machine-readable JSON — structured output ready for n8n, Make, Python, JavaScript, AI Agents
• Source provenance — every price has a source URL, check timestamp, and confidence level

Use cases

Use case	What you get
AI cost optimization	Compare actual API costs across all Chinese providers
Model routing	Input pricing data into your AI Gateway / Model Router
FinOps monitoring	Scheduled runs detect price drops and increases
Free tier discovery	Find models with free quotas for development/testing
Deal tracking	Get notified when your preferred model goes on sale

How to use

1. Run the Actor on Apify Console or via API
2. Choose providers: ["all"] or specific ones like ["deepseek", "qwen"]
3. Select data: toggle pricing, free tiers, deals, and price changes
4. View results in the Output tab or download as JSON/CSV/Excel

Quick start via API

curl -X POST "https://api.apify.com/v2/acts/YOUR_USERNAME~china-ai-pricing-freetier-tracker/runs" \
  -H "Authorization: Bearer <YOUR_APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
    "providers": ["all"],
    "include_pricing": true,
    "include_free_tiers": true,
    "include_deals": true,
    "include_changes": true
  }'

Input

Field	Type	Default	Description
providers	string[]	["all"]	Providers to collect (or "all")
include_pricing	boolean	true	Collect API pricing data (free)
include_free_tiers	boolean	true	Collect free tier/quota info ($1.00/run)
include_deals	boolean	true	Collect promotions and discounts ($0.60/run)
include_changes	boolean	true	Detect price changes since last run ($0.40/run)
only_active_deals	boolean	true	Only include currently active deals

Default run (no input): collects all data types from all providers. Premium charges: $1.00 + $0.60 + $0.40 = $2.00/run. To run free only, set include_free_tiers, include_deals, and include_changes to false.

Output

Datasets

The Actor outputs data to 5 named datasets plus a default dataset:

Dataset	Description	Price
pricing	Current API pricing (original + normalized to USD)	Free
free-tiers	Free quotas and free models	$1.00/run
deals	Active promotions and discounts	$0.60/run
changes	Price changes detected since last run	$0.40/run
Default dataset	Summary of all records (for Console Output tab)	—

All datasets have source fields flattened as scalars: source_type, source_url, source_checked_at, source_confidence, source_effective_from.

Example: Pricing record

{
  "_type": "pricing",
  "provider_id": "deepseek",
  "model_id": "deepseek-v4-flash",
  "original_currency": "CNY",
  "original_unit": "per_1m_tokens",
  "original_input_price": 3.0,
  "original_output_price": 9.0,
  "input_price": 0.414,
  "output_price": 1.242,
  "cached_input_price": 0.0138,
  "batch_input_price": 0.207,
  "batch_output_price": 0.621,
  "currency": "USD",
  "price_unit": "per_1m_tokens",
  "context_window": 1000000,
  "source_type": "official",
  "source_url": "https://api-docs.deepseek.com/zh-cn/quick_start/pricing",
  "source_checked_at": "2026-09-14T10:45:58.589Z",
  "source_confidence": "high"
}

Example: Free tier record

{
  "_type": "free_tier",
  "provider_id": "tencent",
  "model_id": "hunyuan-a13b",
  "free": true,
  "free_type": "permanently_free",
  "free_amount": null,
  "free_unit": null,
  "period": "lifetime",
  "conditions": null,
  "expires_at": null,
  "source_type": "official",
  "source_url": "https://cloud.tencent.com/document/product/1729/97731",
  "source_confidence": "high"
}

Example: Price change record

{
  "_type": "change",
  "provider_id": "deepseek",
  "model_id": "deepseek-v4-flash",
  "change_type": "price_decrease",
  "field": "input_price",
  "old_value": 0.552,
  "new_value": 0.414,
  "change_percent": -25,
  "anomaly": false,
  "detected_at": "2026-09-14T10:00:00Z",
  "source_type": "official",
  "source_url": "actor://change-detector"
}

Key-Value Store outputs

Key	Description
run-summary	Run stats: model count, provider count, duration, per-provider status
provider-status	Per-provider collection result (success/failed, records found, duration)
price-changes-history	All detected changes across all runs
latest-snapshot	Current data snapshot for next-run comparison

Data fields reference

Pricing

Field	Description
original_input_price	Price as shown on the provider's website
original_currency	Original currency (CNY, USD)
original_unit	Original unit (per_1m_tokens, per_1k_tokens)
input_price	Normalized input price (USD per 1M tokens)
output_price	Normalized output price (USD per 1M tokens)
cached_input_price	Cached/buffered input price (if available)
batch_input_price	Batch/off-peak input price (if available)
batch_output_price	Batch/off-peak output price (if available)
source_url	URL where the price was found
source_checked_at	When the price was last checked
source_confidence	Data confidence (high/medium/low)

Free tiers

Field	Description
free_type	Type: permanently_free, limited_time_free, public_beta_free, trial_credit, free_model, token_quota, daily_quota, monthly_quota, new_user_credit, free_requests
free_amount	Amount of free usage (e.g., 1000000 tokens)
free_unit	Unit: tokens, requests
period	Period: daily, monthly, one_time, lifetime
conditions	Conditions: "new users only", "first 14 days", etc.
region_limit	Region restriction (if any)
rate_limit	Rate limit (e.g., "QPS=2")

Deals

Field	Description
deal_type	Type: discount, free_quota, limited_time_price, batch_discount, campaign
original_price	Price before discount
discount_price	Price during deal
discount_percent	Discount percentage
expires_at	When the deal expires (null if no expiry)

Scheduled monitoring

Set up a cron schedule to track changes automatically:

Frequency	Purpose
Every 6 hours	Track fast-changing prices and new deals
Daily	Standard monitoring for most use cases
Weekly	Low-frequency overview

Use Apify Scheduler or trigger via webhook.

Pricing

PPE charges (per-event)

Dataset	Price	Content
pricing	Free	All model prices (input/output/cached/batch)
free-tiers	$1.00/run	4 types of free models with confidence levels
deals	$0.60/run	Active promotions + Free Tier Deathwatch
changes	$0.40/run	Price changes since last run + anomaly detection

Default run (all enabled): $1.00 + $0.60 + $0.40 = $2.00/run

To run free only: set include_free_tiers, include_deals, and include_changes to false in Input.

Platform costs (paid by you, from your Apify plan)

Mode	Approximate time	Compute Units
All providers	~2-4 minutes	~0.1-0.2 CU
Single provider	~10-30 seconds	~0.01-0.03 CU

On Apify's free tier ($5/month), you can run the free pricing dataset daily (~50 runs/month). Premium datasets require a paid Apify plan.

Tips

• Start with ["all"] to see what data is available, then filter to specific providers
• Enable include_changes to detect price drops — this is the most valuable long-term feature
• Use the changes dataset with webhooks to get Slack/email notifications on price changes
• Check provider-status in KVS to see which providers succeeded or failed
• Use anomaly: true in changes to filter out suspicious >90% changes (likely parsing errors)
• Run free only: set include_free_tiers, include_deals, include_changes all to false

Data quality

• Every record includes source_url — you can always verify the data
• Every record includes source_confidence — high/medium/low based on source type
• Prices that cannot be determined are set to null, never guessed
• Anomalous price changes (>90%) are flagged with anomaly: true
• Single provider failures don't crash the entire Actor
• Providers are run with error isolation — one failure doesn't affect others

FAQ

Q: How accurate is the pricing data? A: Prices come from official provider websites and documentation. Each record includes the source URL and check timestamp. We never guess prices — if a price can't be determined, it's set to null.

Q: How often should I run this Actor? A: Daily is sufficient for most use cases. Run every 6 hours if you need to catch fast-changing promotions.

Q: What happens when a provider website changes? A: Each provider runs in isolation. If one provider's website changes, only that provider fails — all others continue working. Check provider-status in KVS for failure details.

Q: Can I get notified of price changes? A: Yes. Set up an Apify Webhook on the Actor's "Succeeded" event. The changes dataset will contain all detected price changes.

Q: Why are some prices null? A: We set prices to null when the source website doesn't clearly state the price, or when the page can't be parsed. We never guess or infer prices from other models.

Q: What is the default run cost? A: With all defaults (no input), the Actor collects pricing (free) + free tiers ($1.00) + deals ($0.60) + changes ($0.40) = $2.00 per run. To run for free, disable the three premium datasets in Input.

Support

• Apify Discord — community support
• Contact the author for feature requests and custom solutions

---

Built for developers who need to track Chinese AI API costs programmatically.