# 热核算力 (CodexHot) Review OpenAI-compatible GPT API aggregation service for developers, billed from a prepaid balance with vouchers starting at CNY 5. ## Editor's Score: 44/100 Capability scores low at 12 because CodexHot is a thin proxy layer with no documented rate card, no published model list with confirmed availability, and no features beyond basic endpoint routing that any OpenAI-compatible reseller also provides. Ease of use earns 14 because the OpenAI-compatible interface genuinely requires minimal integration effort and setup guides exist, but the absence of a public rate card forces developers to integrate blind. Value lands at 10: the CNY 5 minimum voucher is accessible, but with no published per-model prices and no free tier, buyers cannot compare value against direct OpenAI access or competing resellers before spending money. Delivery scores 8 because the tool's own submitter publicly corrected third-party descriptions that attributed features (automatic failover, team consumption) the service does not document, no user sentiment or uptime data was findable in any search, and the overall evidence base for what the service actually delivers is thin. ## Pros - Drop-in OpenAI SDK compatibility means zero code changes are needed to switch the base URL and API key. - CNY 5 minimum voucher gives developers a genuinely low entry point to test the service before loading larger balances. - Prepaid wallet model prevents surprise overage bills, which is a real advantage for solo developers on a fixed budget. ## Cons - No public per-model CNY rate card is available in the documentation, so developers cannot calculate expected costs before integrating, a specific gap compared to direct OpenAI API access which publishes exact per-token prices. - No free tier or trial period exists, meaning every developer must spend real money before they can evaluate latency, reliability, or output fidelity through this proxy. - The service's own submitter confirmed that claims of 'automatic failover' and 'team consumption' are not established in public docs, leaving the actual reliability architecture undocumented. ## Pricing Free. Pricing reflects what we saw at the time of review (2026-10-08). Confirm current pricing on the official site. Category: AI Coding & Dev Tools (https://aitoolseekers.com/category/ai-coding-dev-tools) Official site: https://www.codexhot.com/ Last verified by a human: 2026-10-08 Reviewed by a human at AI Tool Seekers, a hand-reviewed AI tools directory. Every listed tool is verified by a person before publishing. Scores are editorial opinions from a fixed rubric (capability, ease of use, value, delivery on promise) and cannot be bought; there is no pay-to-rank. Directory: https://aitoolseekers.com This review: https://aitoolseekers.com/tools/codexhot Methodology: https://aitoolseekers.com/how-verification-works ## Verdict CodexHot suits developers in China who need an OpenAI-compatible API endpoint they can top up in CNY without a foreign payment method. Its pricing structure is opaque compared to direct API providers, and the absence of a free tier or trial means you are committing real money before you can evaluate output quality or latency. Developers who can pay OpenAI directly, or who need a richer feature set with documented rate limits and SLAs, should look elsewhere. Best for: China-based developers needing CNY-denominated, prepaid GPT API access. Not for: Developers who want a free trial, transparent per-model pricing tables, or documented uptime guarantees before committing funds. ## Overview 热核算力 (CodexHot) is a Chinese API aggregation service that sits in front of GPT models and exposes an OpenAI-compatible endpoint at api.codexhot.com/v1. Developers point their existing OpenAI SDK or HTTP client at that base URL, swap in a CodexHot API key, and their calls route through the service. The appeal is straightforward: CNY-denominated prepaid billing, which sidesteps the foreign-card friction that blocks many China-based developers from paying OpenAI directly. The billing model is pure pay-as-you-go from a prepaid wallet. You redeem a voucher (minimum CNY 5) at codexhot.com/redeem, and the balance drains as you make API calls. Charges vary by model and by what the service calls usage-billed groups, though the public documentation does not publish a full per-model rate card. The supported endpoints include Chat Completions, Responses, and Messages, covering the main patterns developers use with GPT models. What CodexHot is not: it is not a coding assistant, not an IDE plugin, and not an agent framework. It is a proxy and billing layer. That narrow scope means it competes less with tools like Cursor and more with other API resellers or direct OpenAI access. The value proposition stands or falls entirely on price, reliability, and how smoothly the OpenAI-compatible interface actually behaves in practice, none of which the public documentation quantifies with hard numbers. ## Who should use it CodexHot is a practical fit for individual developers or small teams in China who are building applications on top of GPT models and need a payment path that accepts CNY without a foreign credit card. The CNY 5 minimum voucher makes it low-risk to test the endpoint before committing larger balances, which is useful for hobbyists or students prototyping small projects. Anyone outside China who can pay OpenAI directly should skip this service; the added intermediary layer brings no documented benefit and introduces an extra point of failure. Teams that need SLA guarantees, published uptime figures, or a detailed per-model rate card before signing off on a vendor should also look elsewhere. Qveris, listed in our directory, is worth considering for developers who need a pay-as-you-go API routing layer with access to a broader set of models and data sources beyond GPT. ## FAQ Q: Is there a free trial or free tier? A: No. CodexHot has no free plan and no free trial. The minimum entry point is a CNY 5 voucher redeemed at codexhot.com/redeem. Q: What endpoints does CodexHot support? A: The service supports Chat Completions, Responses, and Messages endpoints via an OpenAI-compatible base URL at api.codexhot.com/v1. Q: How does billing work? A: Usage is deducted from a prepaid balance. You load the balance by redeeming vouchers; charges vary by model and usage group. No subscription fee applies. Q: What are the per-model prices in CNY? A: Specific per-model CNY rates are not published in the publicly available documentation. You should check the official site's current model list and pricing page before integrating, as rates vary by model and group. Q: Does CodexHot offer automatic failover or team features? A: The service's own submitter has stated that automatic failover and team consumption features are not established in the public documentation, so these should not be assumed to be available. Q: Do I need to change my code to use CodexHot? A: Minimal changes are required. You replace the OpenAI base URL with api.codexhot.com/v1 and swap in a CodexHot API key. Any SDK or HTTP client already built for OpenAI should work without further modification.