Tencent Hunyuan Pricing After the 2026 TokenHub Migration

The previous console is historical, not a current route

Tencent’s May 12, 2026 migration notice sets June 30 as the end of new purchases and September 30 as full termination of the former API console. This includes the former Hunyuan route and certain associated services. A 2025 invoice may still contain legacy product codes, but a new October 2026 forecast belongs on TokenHub. Preserve old IDs for audit rather than copying discontinued SKU prices into new purchases. Migration can also require a new account entitlement, API base path, model name and rate-limit configuration.

What the Guangzhou list rate actually says

The TokenHub model-price document, updated October 24, 2026, lists Hy3 online inference in the Guangzhou table at CNY 1 per million input tokens, CNY 4 per million output tokens and CNY 0.25 per million cache-hit tokens. A separate batch line lists Hy3 at CNY 0.5 input, CNY 2 output and CNY 0.125 cache hit. These are different service routes: applying another 50% “batch discount” to already listed batch rates would double-count the reduction. An international account or Singapore deployment needs its corresponding location and billing table, not these Guangzhou amounts.

Compute a transparent illustrative invoice

For a Guangzhou Hy3 online workload with 2 million new input tokens, 1 million cache-hit input tokens and 500,000 output tokens, the posted line items would produce CNY 4.25: 2×1 + 1×0.25 + 0.5×4. That assumes the three quantities are already split into mutually exclusive billing buckets, excluding storage, tools and taxes. The same request distribution under the vendor’s listed Hy3 batch meter would be CNY 2.125, before checking batch eligibility and latency requirements. These figures are examples derived from a dated price table, not a customer invoice.

Different schedules mean different operational outcomes

Tencent documents other models, video token meters, peak/off-peak windows for selected hosted third-party offerings, and explicit regional price columns. For models with a time-of-day schedule, the platform uses the server receipt time in Beijing time, not the customer’s local clock or response completion time. Do not assign those rules to Hy3, whose illustrated line carries no peak tariff. A realistic service budget should log route and region on every event and account for retried requests, caching behavior and timeout policies.

A procurement decision must test the new route

Archive the shutdown notice, the October 24 rate table, the actual console quote and first new invoice before retiring the old provider entry. The marketing name “Hunyuan” describes a family, not a single immutable commercial SKU. Confirm API key ownership, residency constraints, access approval and whether there are reserved-capacity purchases. Tencent’s first-party table supports these listed numbers, but cannot verify a particular buyer’s tax status or contract-specific concessions.

Decision checklist and limitations

Before purchase or deployment, retain the literal model or plan identifier, payer/merchant, country or cloud region, billing currency, exact usage meter, source URL, checked date, effective date if explicitly supported, and any plan commitment or promotional eligibility. Recalculate from the companion price and quota snapshots rather than assuming that a remembered list tariff will remain unchanged. A credit allowance, maximum input length, published token rate and confirmed billed currency answer different questions; never add them as if they were the same quantity.

When comparing another provider or route, use the same input/output distribution and acceptable result definition. Include unsuccessful attempts, optional tools, storage, tax, payment and human review only when those components are actually measured or supported by an explicit source. Vendor marketing and documentation are not independent product tests. This article makes no claim to have personally tested paid endpoints, purchased plans, or checked account-specific taxes or availability.