PackyAPI is an enterprise AI control plane that serves as a unified LLM API gateway, connecting to global foundation models. It offers a single domain, key, and risk controls for observability, scalability, and governance. Key capabilities include real-time scheduling with dynamic switching based on health and latency weighting, unified observability for calls and spend, and intelligent rate limiting. The API is compatible with OpenAI's format and supports endpoints for chat completions, embeddings, reranking, image generation/editing/variations, and audio speech/transcription/translation. It integrates with over 30 model providers such as OpenAI, Claude, Gemini, and others, featuring 99.9% SLA availability and 7 regional points of presence. Use cases include building reliable AI infrastructure for teams with access control, cost transparency, and global routing.
PackyAPI
codex-api.packycode.com
PackyAPI (codex-api.packycode.com) is an OpenAI-compatible API relay for Codex and other models via a single interface.
- Models
- 78 models
- From
- $0.011/M
- Speed
- 69 tok/s
- Updated
- 7/30/2026
- Latency
- 0.00 s
- Created At
- 1/5/2026
- Recharge Rate
- ¥1.00 per $1 quota
Features
Login Methods
Billing
Payment methods
API Endpoints
- Endpoint 1Historical / Unverified
https://www.packyapi.com - Endpoint 2Historical / Unverified
https://api-slb.packyapi.com - Main Site Endpoint
https://www.packyapi.aiRecommended for non-mainland areas
- Endpoint 4Historical / Unverified
https://codex-api.packycode.com - Endpoint 5Historical / Unverified
https://codex-api-slb.packycode.com - Global Endpoint
https://cf.api.fanStable and reliable, recommended for production environments
- Optimize the endpoint
https://slb-v1.api.fanOptimize routing and recommend for latency-sensitive users
Verify ownership to unlock provider management features:
- Edit provider name, content, and links
- Get featured with priority traffic and visibility boost
- Display a verified badge to build user trust
Leaderboard Rankings
Rankings are based on community-submitted tests and periodic health probes. Advisory only, not official data.
About PackyAPI
Health Check
API Benchmarks & Pricing
Compare 106 model rows across audit recency, latest speed tests, throughput, latency, and per-token pricing.
qwen3-vl-flashbailian | $0.011/M | $0.107/M |
claude-opus-5aws-q | $0.214/M | $1.07/M |
claude-opus-5cc-sale | $0.571/M | $2.86/M |
| $1.79/M | $8.93/M | |
claude-opus-5claude-officially | $5.00/M | $25.00/M |
gemini-3.6-flashgemini-officially | $1.50/M | $7.50/M |
gemini-3.5-flash-litegemini-officially | $0.300/M | $2.50/M |
kimi-k3kimi-sale | $1.86/M | $9.29/M |
kimi-k3kimi-officially | $2.71/M | $13.57/M |
gpt-5.6-lunacodex | $0.071/M | $0.429/M |
gpt-5.6-terracodex | $0.179/M | $1.07/M |
gpt-5.6-terraazure-officially | $1.79/M | $10.71/M |
gpt-5.6-solcodex | $0.357/M | $2.14/M |
gpt-5.6-solazure-officially | $3.57/M | $21.43/M |
grok-4.5grok-sale | $0.029/M | $0.086/M |
grok-4.5grok-officially | $0.571/M | $1.71/M |
hy3hunyuan-officially | $0.114/M | $0.457/M |
claude-sonnet-5aws-q | $0.086/M | $0.429/M |
claude-sonnet-5cc-sale | $0.229/M | $1.14/M |
claude-sonnet-5cc-expensive | $0.629/M | $3.14/M |
| $0.714/M | $3.57/M | |
claude-sonnet-5claude-officially | $2.00/M | $10.00/M |
claude-opus-4-1-20250805 | - | - |
gpt-image-2image | $0.057/request | - |
gpt-image-2sora | $0.011/request | - |
Showing 25 of 106 model rows
Recent Test Records
| Time | Model | Speed | Latency |
|---|---|---|---|
| Jun 17, 06:36 AM | claude-sonnet-4-6 | 42.12 tok/s | 4.50s |
| May 27, 03:52 AM | claude-opus-4-7 | 45.94 tok/s | 2.21s |
| Jan 5, 02:13 AM | gpt-5.2 | 111.67 tok/s | 9.52s |
| Jan 5, 02:12 AM | gpt-5.2 | 75.76 tok/s | 13.46s |
Similar API Provider Alternatives to Compare
Compare PackyAPI alternatives against 6 nearby API providers using 401 LMSpeed signals across shared model coverage, pricing, benchmark speed, uptime, and free-model availability.
| Provider | Why compare | Models | Free | Avg price | Speed | 30d uptime |
|---|---|---|---|---|---|---|
| PackyAPI codex-api-packycode-com PackyAPI (codex-api.packycode.com) is an OpenAI-compatible API relay for Codex and other models via a single interface. | Current provider baseline | 78 | 0 | $0.041/M | 69 tok/s | 9970% |
| V-API v-api Provides access to over 500 AI models via OpenAI protocol with pay-as-you-go pricing, fast response times, and transparent billing. |
| 699 | 15 | $0.010/M | 130 tok/s | 9950% |
| C85 API c85-api C85 API is an OpenAI-compatible API gateway built on the New API panel, providing unified access to multiple AI models through a hosted endpoint. |
| 188 | 130 | N/A | N/A | 0% |
| NVIDIA NIM nvidia-nim NVIDIA NIM provides optimized AI model inference APIs for LLMs, vision, and embedding models through NVIDIA cloud infrastructure. |
| 166 | 0 | N/A | 78 tok/s | 9960% |
| Privnode privnode Privnode appears to provide an OpenAI-compatible API endpoint on privnode.com. The public root is protected by a Cloudflare challenge during review. |
| 92 | 28 | N/A | 36 tok/s | 0% |
| Elysiver API elysiver-api Elysiver API appears to provide an OpenAI-compatible API gateway at elysiver.h-e.top. The public root is protected by a Cloudflare challenge during review. |
| 78 | 0 | $0.014/M | 119 tok/s | 9960% |
| GPT Load (Shiho) gpt-load-shiho-top GPT Load (Shiho) is an OpenAI-compatible API load balancing service hosted at gpt-load.shiho.top, distributing requests across multiple AI model providers for improved reliability. |
| 27 | 0 | N/A | 1164 tok/s | 9980% |
Announcements
To improve access speed and request stability for global users, we have redesigned and optimized the Endpoint architecture.
This Endpoint change does not affect the console website
After the adjustment, the following Endpoints will serve global requests:
The main sites www.packyapi.com and www.packyapi.ai will apply regional restrictions to model calls.
The transition period is from today to August 10 (CST), please complete the Endpoint switch as soon as possible to avoid impact on subsequent calls.
Thank you for your understanding and support.
PackyAPI Team
OpenAI service is unstable today, you may experience frequent disconnections, retries, cache anomalies, and server-side errors.
Users who require high stability are recommended to temporarily switch to other model groups.
If you need to continue using GPT series models, you can migrate to the Azure group for a more stable experience.
PackyAPI Team
Grok-Officially group will adjust the multiplier to 1.5X starting today, with a promotional event lasting one week, during which related calls will be billed at the adjusted multiplier.
After the event, the original multiplier will be restored. Users with needs are advised to arrange usage time accordingly.
PackyAPI Team
Effective immediately, the GLM-5.3 zai-off group and glm-sale group are now offering a 50% limited-time discount. The end date of the promotion will be announced separately.
During the promotion period, billing rates are calculated at 50% of the original price, applicable to all requests under these groups.
For details or to adjust your usage plan, please contact customer service or follow future announcements.
PackyAPI Team
Codex group has launched the gpt-6-astra model, with a billing multiplier of 0.5x. Welcome to use it.
Thank you for your understanding and support!
PackyAPI Team
deepseek-sale group has officially launched the deepseek-v4-pro model, which can now be called via this group. For specific billing multipliers and detailed parameters, please refer to the platform's relevant instructions.
PackyAPI Team
DeepSeek-V4-Flash model has officially launched in the DeepSeek-Sale group.
DeepSeek-Sale group starts a limited-time promotion from today, with a billing multiplier of 0.25X, lasting one week.
For details, please contact customer service. We welcome everyone to use it actively.
PackyAPI Team
Kimi-Sale group is currently undergoing maintenance, expected to be restored in half a day. Upon restoration, the promotion will end and the group multiplier will return from 0.3 to 0.5.
During maintenance, this group may not be usable normally. It is recommended to temporarily use other groups for related tasks. We apologize for any inconvenience caused.
PackyAPI Team
claude-fable-5-1 model has been officially launched in the CC group, with a billing multiplier of 2x.
Subsequently, it will be gradually extended to other groups based on usage. We welcome you to try it and provide feedback.
Thank you for your understanding and support!
PackyAPI Team
Grok-Sale is expected to resume service today.
After service restoration, the service multiplier is expected to revert from 0.1x to 0.3x on Thursday.
During the adjustment period, please monitor service status and billing changes, and arrange usage accordingly.
PackyAPI Team
From today until September 7, 2026, the billing multiplier for the codex-cyber group is adjusted from 1x to 0.6x, lasting one week.
During the promotion, the usage cost of this group is reduced; feel free to use it as needed.
PackyAPI Team
Codex has launched a tiered billing model; requests with context exceeding 272K will be billed at tiered rates.
At the same time, it is reiterated that any form of Cyber-related violations discovered three times cumulatively will result in direct account suspension. Please strictly adhere to the platform usage guidelines.
PackyAPI Team
Effective immediately, Cyber-related violations will no longer be eligible for exemption. Please strictly follow the usage guidelines to avoid account suspension due to Cyber violations.
Blue team users can continue related work using the codex-cyber group, while the red team is not supported for now.
We kindly ask all users to comply with regulations and jointly maintain a good platform order.
PackyAPI Team
The codex-cyber group is now officially live, with a billing multiplier of 1x, and includes gpt-daybreak-blue-latest model access.
This group has limited resources and is for compliant use only; any third-party calls or resale are strictly prohibited.
PackyAPI Team
The Sora group has been officially discontinued, and related requests will no longer be processed.
For future image generation tasks, please use the Image group for API calls.
We apologize for any inconvenience caused.
PackyAPI Team
DeepSeek Official Group has added the DeepSeek-V4-Flash-Vision-Exp model, now available for API calls with a limited-time 20% discount.
For specific billing rates and promotion period, please refer to the platform billing page.
PackyAPI Team
The Grok-Sale, Aws-Q, and CC-Sale groups are currently undergoing service restoration. During this period, you may experience unstable connections, response delays, or temporary unavailability. We recommend waiting patiently or temporarily using other groups.
Restoration progress will be announced separately. We apologize for any inconvenience caused.
PackyAPI Team
Due to upstream policy adjustments, the Codex group is currently running in degraded mode, and some request latencies may increase.
At the same time, request clients are currently restricted. It is recommended to use first-party OpenAI programs such as Codex for calls. Other clients may experience request failures or rejections.
The above adjustments are temporary measures and will be restored or adjusted in a timely manner based on the upstream situation.
PackyAPI Team August 2026
Notes
- Health checks: Scope: the 72-hour chart and recent availability measure API connectivity only. Each bar summarizes one hour of checks. Targets: LMSpeed tries the configured health check URL and provider status URL first, then API endpoints derived from known API hosts and recent speed-test base URLs. A website host is considered only when it looks like an API endpoint. Probe steps: each candidate goes through DNS lookup, TCP connection, TLS handshake for HTTPS, and an HTTP HEAD request with redirects followed. Probing stops after the first reachable candidate. Reachable criteria: every required network step must succeed. An HTTP response below 500 is treated as reachable, including 401 because it confirms that an authenticated API endpoint responded, except for statuses classified as blocked. Blocked results: HTTP 403, 429, 521, 525, and 530, plus detected WAF or Cloudflare challenges, are shown as blocked and excluded from availability calculations because LMSpeed cannot determine whether the API itself is down. Model availability: when a dedicated test key is configured, LMSpeed sends an authenticated GET request to a derived /models endpoint and compares returned model IDs with this provider's listed models. These per-model results appear in Models & Pricing and are not included in the provider connectivity percentage. Timeouts: TCP connection, TLS handshake, HTTP connectivity, and model requests each use a 20-second timeout. A full run can take longer when several candidates are tried. Frequency: a background worker checks all providers every 5 minutes by default. The 72-hour chart combines those samples into hourly bars, and the schedule may be changed by the service operator. Limit: automated samples are not an SLA and do not guarantee account quota, every model, every region, or successful completion requests. Check the provider's own status page before making operational decisions.
- Domain Rating data is sourced from Ahrefs. It is a 0–100 backlink-based domain strength signal and does not measure API speed or reliability.
- Announcements and FAQ are read from this provider's NewAPI status snapshot when available. LMSpeed stores the original content and optional English translations from the provider status source, then shows the localized fields on this page.