IXCDN: Free CDN + Agent Hosting

Free CDN + Agent Hosting

IXCDN runs a free CDN and agent hosting for global websites, AI apps and cross-border workloads. Edge nodes cache static content, route dynamic requests by region and carrier, and enforce DDoS, WAF and TLS at the ingress. CDN traffic, AI tokens, storage, compute and enterprise routes are metered separately. Live in 3 minutes, no minimums, no prepayment.

Start for Free Request a Custom Plan
  • Instance Active
  • Free Forever
  • No Credit Cards
  • Usage Billing

Global Acceleration

Global edge cache and path-aware origin fetch for websites, APIs, downloads and media.

Model Gateway

One OpenAI-compatible ingress for model routing, keys, rate limits and quotas.

Website Builder

Upload static files or build output, bind your domain, publish sites and docs.

Edge Security

DDoS scrubbing, WAF rules and access control at the edge, in front of origins and APIs.
IXCDN + Edge

Start free Accelerate globally

Add a domain or AI API and complete DNS verification. Requests then go to the nearest edge node, which applies cache, security and routing rules and fetches from origin only when needed.

Workloads Websites and APIsDownload deliveryLLM / AIGC streamingGaming and interactive workloads
IXCDN Edge GSLB + Cache + WAF
Edge capabilities DNS routingEdge cacheOrigin shieldAI ingressWAF security+ more services

Line routing

Routing by geography, carrier, ASN and recent path quality cuts detours and needless origin fetches.

Security at the edge

DDoS, WAF, bot management and token validation run at the edge, so bad requests do not enter origin.

Cache rules

Separate cache rules for static assets, large files and media raise edge hit ratio.

Usage and billing

Traffic, token usage, security events and billing share one view, so ops and finance read the same numbers.

CDN Websites, APIs, downloads and media
AI Token streams and model APIs
91% Edge cache hit ratio
View capabilities Read the blog

AI Gateway

One ingress for LLM APIs

An edge AI gateway for LLM apps and agents: multi-model access, streamed responses, usage accounting and key isolation.
OpenAI-compatible API A single request format; most application code stays as is when you add a model.
Streaming Nearby ingress and reused long connections cut first-token wait by 38%.
Usage and cost Token and request counts by model, workload and endpoint, allocatable to projects.
Keys and access Upstream keys stay in the gateway, scoped by team, app and environment with rate limits.
Failover Route across models by policy, with switchover or fallback when an upstream fails.
Logs and troubleshooting Request and error logs by app, model and status code for tracing upstream issues and cost anomalies.

Capabilities

One edge network for global delivery

Performance

Serve remote users from the nearest node

The farther a user is from origin, the more latency and packet loss you carry. IXCDN caches static content on edge nodes and routes dynamic requests by path quality, so pages, APIs, downloads and model streams come back from closer. Onboard one workload and compare latency before and after.

17:24 IXCDN Edge 17.8ms Provider A 18.9ms Provider B 20.4ms Provider C 20.8ms Provider D 22.1ms
1 IXCDN Edge 17.8ms
2 Provider A 18.9ms
3 Provider B 20.4ms
4 Provider C 20.8ms

Routing

Less detour, fewer origin fetches, a bill you can read

Ingress is chosen by region, carrier, ASN and node load, which cuts cross-network detours and repeated origin fetches. CDN traffic, AI tokens, storage, compute and enterprise routes sit on separate meters so budgets map to projects.

User IXCDN Origin SG TYO
Path qualityBest
Origin fetches-42%
Routing policyLive

Security

The acceleration ingress is also the security ingress

DDoS scrubbing, WAF, TLS, token validation and access control run before traffic reaches origin or model services. Start in observe mode with log-only rules, check false positives, then move to intercept and ban.

TLS Passed
WAF Observed
Bot Blocked
Token Verified

Console

Onboarding, monitoring and scale in one console

Domains, certificates, cache, security, logs, AI Gateway and billing live together. Start with small traffic, check hit ratio and error rate, then migrate more sites, APIs and model traffic.

IXCDN Control Centre Ready

Validate with one onboarding flow

DNS validationDone
HTTPS certificateReady
Cache policyOn
Access logsLive
CAP-01

Lower cross-border latency

Edge nodes answer nearby, static content is cached and dynamic requests route by region, carrier and real-time quality. Websites, APIs, downloads and model streams take fewer round trips.

CAP-02

Lower origin and bandwidth cost

Up to 91% edge cache hit ratio, origin fetch collapsing and duplicate-request filtering cut origin bandwidth and server load. Traffic spend maps to specific domains and paths.

CAP-03

Layered protection at the ingress

DDoS scrubbing, WAF rules, bot protection and API authentication run before requests reach origin, covering L3 to L7 and common audit requirements.

CAP-04

Usage and cost split by line item

Access logs, live monitoring and billing itemize traffic, AI tokens and compute. Anomalies trace to URL, status code and region, giving ops and finance the same numbers.

Features and limits

Performance, security, cache, access control, streaming and developer APIs, grouped so teams can check IXCDN against each workload.

  • IPv4 / IPv6:

    Edge nodes, origin links and IX interconnects are dual-stack, so clients can connect over IPv4 or IPv6.

  • HTTP/2:

    Multiplexes requests on one connection, cutting handshakes and head-of-line blocking for asset-heavy pages.

  • HTTP/3:

    Runs over QUIC and UDP, so reconnects are faster and packet loss hurts less on mobile and weak networks.

  • TLS 1.3 with 0-RTT:

    TLS 1.3 removes a handshake round trip; 0-RTT resumes sessions for returning clients to cut reconnect latency.

  • Anycast and GeoDNS routing:

    One address is announced from multiple nodes; GeoDNS, live latency, carrier and line data steer users to a nearby ingress.

  • Dynamic content acceleration:

    When a dynamic request misses cache, connection reuse, route selection and origin link optimization reduce round trips.

  • Edge rules:

    Run redirects, URL rewrites, header changes and access checks at the edge, configured per path or condition.

  • Extended API:

    Open APIs for domains, cache, certificates and logs, so IXCDN fits your release and operations tooling.

  • Terraform support:

    Declare domains, cache and security in Terraform, with changes reviewed and version-controlled.

  • Grafana support:

    Feed metrics into Grafana alongside your existing dashboards and alert rules.

  • Origin shielding:

    A convergence layer in front of origin merges requests from many edge nodes before they hit origin.

  • Origin groups:

    Configure multiple origins with primary/backup, weighted or health-checked distribution, with automatic failover.

  • Purge by URL, file, directory:

    Submit a purge by full URL, file name or directory and have edge nodes invalidate within seconds.

  • Prefetch:

    Pull resources to the edge before users ask, so first visits can hit cache.

  • Cache sharding:

    Split large objects across nodes so per-node capacity and read pressure stay manageable.

  • Always online / Serve stale:

    Serve existing cache when origin times out or is unreachable, then refresh from origin after recovery.

  • Customizable TTL:

    Set cache lifetimes by path, file type or status code to balance freshness and origin fetch volume.

  • CDN cache purge:

    Invalidate specified content across all edge nodes in one submission instead of node by node.

  • Origin validation using SSL:

    Validate certificate and hostname on origin fetch so requests are not steered to a fake origin.

  • Secure token:

    Issue signed, expiring tokens and reject requests with a bad signature or expired token.

  • Access control list:

    Allow or deny by IP, CIDR or region, with rules bound to a domain or path.

  • CORS headers support:

    Add Access-Control-* response headers at the edge so frontends can call cross-origin without origin changes.

  • Protect the origin-CDN connection:

    Force HTTPS on origin fetch, validate certificates and limit which nodes may fetch, reducing man-in-the-middle risk.

  • Brotli compression:

    Compress HTML, CSS and JS with Brotli, usually a little smaller than gzip for text.

  • Rewrite option:

    Rewrite request paths at the edge for legacy links, directory changes and multi-site routing.

  • WebP and AVIF compression:

    Convert to WebP or AVIF based on the Accept header, cutting image bytes while keeping quality.

  • Image quality settings:

    Set compression quality per context and trade sharpness against bandwidth.

  • Image cropping:

    Crop a region by parameter for lists, covers and other display sizes.

  • Image resizing:

    Return dimensions matched to device width and DPR, reducing mobile download volume.

  • Large file delivery optimization:

    Tune chunking, concurrency and timeouts for installers and video to reduce aborts and failures.

  • WebSockets support:

    Pass through Upgrade and long-lived connections for chat, market data, consoles and multiplayer.

  • Custom HTTP status code:

    Return a chosen status code for canary releases, maintenance pages and client handling.

  • Download speed limit:

    Cap download rate per connection or per IP so one user cannot consume all bandwidth.

  • Detailed statistics:

    Review bandwidth, requests, cache hit ratio, status codes and region, filtered by domain and time window.

  • Raw logs:

    Export raw access logs with time, IP, URL, status, cache status and duration for audit and troubleshooting.

  • Log viewer:

    Search and filter logs in the console instead of pulling them down first.

  • Control of consumed traffic:

    Set traffic and request thresholds, alert or throttle past them, and find abnormal traffic and hot URLs.

  • Real Time Monitor:

    Watch traffic, hit ratio and error rate by the minute, then push rate limits, redirects or blocks.

  • SSL/TLS encryption:

    Enable TLS between clients and the edge, with free certificates, your own certificates and forced HTTPS.

  • L3/L4 DDoS protection:

    Scrub volumetric, protocol and reflection attacks at the edge; attack traffic is dropped and not billed.

  • L7 DDoS protection:

    Detect high request rates, slow connections and application patterns, easing attacks on endpoints and pages.

  • WAF:

    Match requests against OWASP rules in observe, intercept, or intercept-and-ban modes.

  • API security:

    Protect APIs with authentication, rate limits, tokens and access control, separating normal calls from abuse.

  • Bot protection:

    Identify scrapers, credential stuffing and automation by behavior and fingerprint, applied to login, signup and query endpoints.

  • Reduced TTFB:

    Edge cache and nearby origin fetch shorten time to first byte for pages and media start.

  • RAM caching for live:

    Keep recent segments in memory to cut origin reads and start time for low-latency live.

  • Live HTTP-based protocols:

    Support HLS and DASH with segment delivery for adaptive-bitrate playback.

  • Live non-HTTP protocols:

    Support RTMP ingest and WebRTC for low-latency interactive live.

  • HESP protocol:

    Deliver low-latency live streams over HESP to shorten end-to-end delay.

  • Micro-segmented caching:

    Cache media in small segments so start and recovery from stalls are faster.

  • Restricted access:

    Limit media access with tokens, Referer, IP and region rules to meet licensing and rights requirements.

  • Comprehensive Documentation:

    API docs cover endpoints, request and response fields, error codes and samples.

  • Authentication and Authorization:

    Control call permissions with API keys or OAuth, scoped by role for reads and writes.

  • Real-time Analytics:

    Get request volume, hit ratio, bandwidth and region by the minute for capacity and anomaly checks.

  • Edge Function Deployment:

    Run custom JavaScript at the edge for auth, rewrites and canary logic without an origin round trip.

  • IPv4 / IPv6:

    Edge nodes, origin links and IX interconnects are dual-stack, so clients can connect over IPv4 or IPv6.

  • HTTP/2:

    Multiplexes requests on one connection, cutting handshakes and head-of-line blocking for asset-heavy pages.

  • HTTP/3:

    Runs over QUIC and UDP, so reconnects are faster and packet loss hurts less on mobile and weak networks.

  • TLS 1.3 with 0-RTT:

    TLS 1.3 removes a handshake round trip; 0-RTT resumes sessions for returning clients to cut reconnect latency.

  • Anycast and GeoDNS routing:

    One address is announced from multiple nodes; GeoDNS, live latency, carrier and line data steer users to a nearby ingress.

  • Dynamic content acceleration:

    When a dynamic request misses cache, connection reuse, route selection and origin link optimization reduce round trips.

  • Edge rules:

    Run redirects, URL rewrites, header changes and access checks at the edge, configured per path or condition.

  • Extended API:

    Open APIs for domains, cache, certificates and logs, so IXCDN fits your release and operations tooling.

  • Terraform support:

    Declare domains, cache and security in Terraform, with changes reviewed and version-controlled.

  • Grafana support:

    Feed metrics into Grafana alongside your existing dashboards and alert rules.

  • Origin shielding:

    A convergence layer in front of origin merges requests from many edge nodes before they hit origin.

  • Origin groups:

    Configure multiple origins with primary/backup, weighted or health-checked distribution, with automatic failover.

  • Purge by URL, file, directory:

    Submit a purge by full URL, file name or directory and have edge nodes invalidate within seconds.

  • Prefetch:

    Pull resources to the edge before users ask, so first visits can hit cache.

  • Cache sharding:

    Split large objects across nodes so per-node capacity and read pressure stay manageable.

  • Always online / Serve stale:

    Serve existing cache when origin times out or is unreachable, then refresh from origin after recovery.

  • Customizable TTL:

    Set cache lifetimes by path, file type or status code to balance freshness and origin fetch volume.

  • CDN cache purge:

    Invalidate specified content across all edge nodes in one submission instead of node by node.

  • Origin validation using SSL:

    Validate certificate and hostname on origin fetch so requests are not steered to a fake origin.

  • Secure token:

    Issue signed, expiring tokens and reject requests with a bad signature or expired token.

  • Access control list:

    Allow or deny by IP, CIDR or region, with rules bound to a domain or path.

  • CORS headers support:

    Add Access-Control-* response headers at the edge so frontends can call cross-origin without origin changes.

  • Protect the origin-CDN connection:

    Force HTTPS on origin fetch, validate certificates and limit which nodes may fetch, reducing man-in-the-middle risk.

  • Brotli compression:

    Compress HTML, CSS and JS with Brotli, usually a little smaller than gzip for text.

  • Rewrite option:

    Rewrite request paths at the edge for legacy links, directory changes and multi-site routing.

  • WebP and AVIF compression:

    Convert to WebP or AVIF based on the Accept header, cutting image bytes while keeping quality.

  • Image quality settings:

    Set compression quality per context and trade sharpness against bandwidth.

  • Image cropping:

    Crop a region by parameter for lists, covers and other display sizes.

  • Image resizing:

    Return dimensions matched to device width and DPR, reducing mobile download volume.

  • Large file delivery optimization:

    Tune chunking, concurrency and timeouts for installers and video to reduce aborts and failures.

  • WebSockets support:

    Pass through Upgrade and long-lived connections for chat, market data, consoles and multiplayer.

  • Custom HTTP status code:

    Return a chosen status code for canary releases, maintenance pages and client handling.

  • Download speed limit:

    Cap download rate per connection or per IP so one user cannot consume all bandwidth.

  • Detailed statistics:

    Review bandwidth, requests, cache hit ratio, status codes and region, filtered by domain and time window.

  • Raw logs:

    Export raw access logs with time, IP, URL, status, cache status and duration for audit and troubleshooting.

  • Log viewer:

    Search and filter logs in the console instead of pulling them down first.

  • Control of consumed traffic:

    Set traffic and request thresholds, alert or throttle past them, and find abnormal traffic and hot URLs.

  • Real Time Monitor:

    Watch traffic, hit ratio and error rate by the minute, then push rate limits, redirects or blocks.

  • SSL/TLS encryption:

    Enable TLS between clients and the edge, with free certificates, your own certificates and forced HTTPS.

  • L3/L4 DDoS protection:

    Scrub volumetric, protocol and reflection attacks at the edge; attack traffic is dropped and not billed.

  • L7 DDoS protection:

    Detect high request rates, slow connections and application patterns, easing attacks on endpoints and pages.

  • WAF:

    Match requests against OWASP rules in observe, intercept, or intercept-and-ban modes.

  • API security:

    Protect APIs with authentication, rate limits, tokens and access control, separating normal calls from abuse.

  • Bot protection:

    Identify scrapers, credential stuffing and automation by behavior and fingerprint, applied to login, signup and query endpoints.

  • Reduced TTFB:

    Edge cache and nearby origin fetch shorten time to first byte for pages and media start.

  • RAM caching for live:

    Keep recent segments in memory to cut origin reads and start time for low-latency live.

  • Live HTTP-based protocols:

    Support HLS and DASH with segment delivery for adaptive-bitrate playback.

  • Live non-HTTP protocols:

    Support RTMP ingest and WebRTC for low-latency interactive live.

  • HESP protocol:

    Deliver low-latency live streams over HESP to shorten end-to-end delay.

  • Micro-segmented caching:

    Cache media in small segments so start and recovery from stalls are faster.

  • Restricted access:

    Limit media access with tokens, Referer, IP and region rules to meet licensing and rights requirements.

  • Comprehensive Documentation:

    API docs cover endpoints, request and response fields, error codes and samples.

  • Authentication and Authorization:

    Control call permissions with API keys or OAuth, scoped by role for reads and writes.

  • Real-time Analytics:

    Get request volume, hit ratio, bandwidth and region by the minute for capacity and anomaly checks.

  • Edge Function Deployment:

    Run custom JavaScript at the edge for auth, rewrites and canary logic without an origin round trip.

Solutions

How IXCDN handles each workload

Global acceleration

Global Website & API Acceleration

Pain
Cross-border routes jitter, so overseas users see high latency; multi-region deployment means maintaining origins and sync yourself.
Fix
Static assets cached on global edge nodes, dynamic requests routed by path quality, HTTPS end to end.
Gain
80%+ faster page loads overseas, 60% lower origin bandwidth cost, one console for the global footprint.
Start free
Fewer origin fetches

Media, Downloads & Large Files

Pain
Installers and video download slowly with high failure rates; hot content saturates origin bandwidth.
Fix
Edge prefetch, chunked caching and download throttling deliver large files from nearby nodes.
Gain
99.9% download success, 90% less origin load, steady delivery for terabyte-scale files.
Start free
Token stream acceleration

AI Apps & Token Streaming

Pain
LLM API latency varies and streaming stalls; keys are scattered and usage numbers don't reconcile.
Fix
An edge AI gateway accepts traffic nearby, reuses long-lived connections and fronts multiple models behind one API.
Gain
38% lower first-token wait for smoother chat; token usage metered by model and workload.
Start free
Interactive ingress

Gaming, Interactive & Realtime

Pain
Update packages distribute slowly and campaign spikes saturate origin; realtime APIs and long connections jitter.
Fix
Nearby ingress, carrier-aware routing and WebSocket pass-through, with security policy applied at the edge.
Gain
Update packages and campaign pages open faster, spikes are absorbed, interactive latency stays steadier.
Start free
Origin-front protection

Enterprise Security Edge

Pain
Origins sit directly on the public internet and face DDoS, scraping and API attacks; audit suffers from incomplete logs.
Fix
All traffic enters the edge first, with layered protection up front and access logs retained for audit.
Gain
99.9%+ attack blocking, origin no longer exposed, meeting baseline security and data compliance.
Start free
Cost and audit

Logs, Billing & Cost Operations

Pain
Traffic, tokens and compute land on one bill, so usage won't reconcile and anomalies are hard to trace.
Fix
Access logs, live monitoring and split billing meter CDN, AI and compute separately.
Gain
Usage, cost and anomalies in one place, shared numbers for ops and finance, and data for scaling decisions.
Start free

Pricing

Usage-based pricing, published rates

The Free plan is free forever. Upgrade when you need steadier global routes, mainland China optimization or DDoS scrubbing.

Free plan

Free Plan

Basic global connectivity, edge cache, universal SSL and access logs, within plan traffic and request quotas. For personal sites, test environments and light workloads.

— Priced by resources; contact sales
Contact sales

What's included

  • Loading plan capabilities…
See all capabilities and quotas
Billed by delivery

Pro Global

Steadier international routes, low-latency routing and origin optimization, billed by delivered traffic. For production websites, APIs and downloads.

— Priced by resources; contact sales
Contact sales

What's included

  • Loading plan capabilities…
See all capabilities and quotas
China + DDoS

China Optimized Shield

Mainland China access optimization, DDoS scrubbing and stronger edge security, billed by protection and route resources. For IPv4 workloads with high protection needs.

— Priced by resources; contact sales
Contact sales

What's included

  • Loading plan capabilities…
See all capabilities and quotas
Enterprise contract

Enterprise Lines

Custom contract covering regions, bandwidth, nodes, scrubbing capacity and SLA. For key-city access and dedicated network resources.

— Priced by resources; contact sales
Contact sales

What's included

  • Loading plan capabilities…
See all capabilities and quotas

Prices and quotas follow the public plan matrix; AI tokens, storage, compute and enterprise routes are metered separately.

View full pricing and feature comparison →

Operational proof

Every request, block and charge is traceable

Cost CDN traffic, AI tokens, storage, compute and enterprise routes metered separately
Latency Edge cache, connection reuse and path-aware routing cut page, API and first-token waits
Security DDoS, WAF, TLS, token validation and origin protection run at the ingress

FAQ

Four questions before onboarding

Which workloads fit IXCDN?

Websites, SaaS consoles, public APIs, installer downloads, media, interactive workloads and AI streaming endpoints. The Free plan is enough to start.

What counts as billable traffic?

Only traffic delivered to end users is billed. Peer networks, optimized routes and enterprise capabilities are separate line items, not folded into one rate.

Where does security run?

DDoS, WAF, TLS, token validation, access control and origin protection run at the edge ingress. Scrubbing capacity, rulesets, concurrency and enterprise policy follow the plan or contract.

What is Mainland China and TOP CBD optimization?

TOP CBD routes traffic to key-city business districts in mainland China for cross-border businesses that need steadier access there. Results depend on local networks, compliance and available routes.

Start with IXCDN

Start with one domain

Live in 3 minutes. The Free plan has no minimums.
Cost CDN traffic, AI tokens, storage, compute and enterprise routes metered separately
Latency Edge cache, connection reuse and path-aware routing cut page, API and first-token waits
Security DDoS, WAF, TLS, token validation and origin protection run at the ingress

Blog

Latest from the blog

All posts