Back to all plans
腾讯云
Hy3 preview
Personal multi-model + enterprise team quotas, from ¥28/mo
Token PlanToken Plan
SubscriptionToken API
Core models
AutoMiniMax-M2.5MiniMax-M2.7MiniMax-M3GLM-5GLM-5.1GLM-5-TurboKimi-K2.5Tencent HY 2.0 InstructTencent HY 2.0 ThinkHunyuan-T1Hunyuan-TurboSDeepSeek-V4-FlashDeepSeek-V4-ProDeepSeek-V4-Flash 原厂直供DeepSeek-V4-Pro 原厂直供Kimi-K2.6Hy3 preview
Auto
Auto routing—personal General uses tc-code-latest, enterprise uses auto; Lite enterprise is Auto-only; also on personal General and enterprise Pro.
MiniMax-M2.5
minimax-m2.5—personal General and enterprise Pro.
MiniMax-M2.7
minimax-m2.7—personal General and enterprise Pro.
MiniMax-M3
minimax-m3—enterprise Pro only, not on personal plans.
GLM-5
glm-5 / glm-5-0—personal General and enterprise Pro.
See the official site for more models
Additional core model names still appear above, with full details on the latest official page.
Plan details
Usage
¥39/mo includes 35M tokens for first-time OpenClaw-style agent trials—officially ~70 Q&A-style rounds as a baseline estimate, varying by length and task complexity.
Models
Supports Auto routing plus MiniMax-M2.5/M2.7, GLM-5/5.1, and Kimi-K2.5—cache-hit/miss input and output all deduct from one pool.
Highlights
Suited to web actions, file handling, data analysis, and scheduled tasks at entry level—coding/agent tools only, no API scripts or backends.
Shares one API key with Hy Token Plan but separate quotas—each account may hold one General and one Hy plan at most.
Shares one API key with Hy Token Plan but separate quotas—each account may hold one General and one Hy plan at most.
Best for
Entry users trying OpenClaw-style agents and multi-model Token Plan for the first time
Usage
¥99/mo includes 100M tokens—roughly 2.9× Lite—with an official ~200-round estimate for daily office work and light development.
Models
Same General model pool and Auto routing—Tencent states 50%+ savings vs direct text-generation API for equivalent usage.
Highlights
Suited to batch file processing, demos, and automation workflows—upgrade supported, no downgrade or refund.
Concurrency Standard > Lite with dynamic platform tuning—account sharing is prohibited and may limit benefits.
Concurrency Standard > Lite with dynamic platform tuning—account sharing is prohibited and may limit benefits.
Best for
Daily office work, light development, and medium-intensity AI coding users
Usage
¥299/mo includes 320M tokens—about 3.2× Standard—for developers and power users who rely on AI daily.
Models
Suited to multi-repo work, complex logic, refactors, and agent orchestration—still the full General multi-model pool.
Highlights
Concurrency Pro > Standard > Lite—if not renewed, unused tokens expire and keys stop working immediately in dependent tools.
For cheaper in-house Hy3, buy Hy line separately—quotas and billing months are independent per series.
For cheaper in-house Hy3, buy Hy line separately—quotas and billing months are independent per series.
Best for
High-frequency AI developers, complex agent orchestration, and users treating Token Plan as a primary workbench
Usage
¥599/mo includes 650M tokens—the top General tier for users who treat AI as core productivity.
Models
Suited to full-stack AI generation, multi-agent collaboration, and CI automation—Max has the highest concurrency among General tiers.
Highlights
Still interactive tools only—no backend API; budget long-term at the ¥599 list price.
Teams needing credits, multi-key quotas, and larger monthly budgets should see Token Plan Enterprise Pro (product/1823/130659).
Teams needing credits, multi-key quotas, and larger monthly budgets should see Token Plan Enterprise Pro (product/1823/130659).
Best for
Heavy AI developers, multi-agent collaboration, and sustained high-pressure coding users
Usage
¥28/mo Hy entry with 35M tokens—Hy3 preview in-house model only, cheaper than General Lite (¥39).
Models
Hy3 preview is agent-oriented—295B/21B MoE, 256K context; no image/video multimodal yet.
Highlights
Shares API key with General plan—Hy pool deducts by Model ID; quotas are independent.
Deal
For users focused on Tencent Hy without GLM/Kimi switching—officially ~70 baseline interaction rounds.
Best for
Users wanting a lower-barrier trial of in-house Hy3 preview
Usage
¥78/mo with 100M tokens—Hy daily tier, Hy3 preview only—official ~200-round estimate.
Models
Hy line is more economical than General for equivalent usage—suited to doc automation, coding agents, and multi-step tool calls.
Highlights
No downgrade or refund—validity is calendar months from purchase (e.g. Jan 31 purchase ends Feb 28 9:59:59).
For third-party model comparison, buy General Token Plan separately—both series can be held together.
For third-party model comparison, buy General Token Plan separately—both series can be held together.
Best for
Users relying on Hy3 daily for agents and document automation
Usage
¥238/mo with 320M tokens—about 3.2× Hy Standard—for daily high-frequency Hy3 development.
Models
Suited to complex refactors, long-document analysis, and multi-turn agents—Hy Pro list (¥238) undercuts General Pro (¥299).
Highlights
Concurrency Hy Pro > Hy Standard > Hy Lite—platform adjusts rate limits dynamically.
Personal Token Plan issues one shared API key—Hy tiers cannot downgrade once purchased.
Personal Token Plan issues one shared API key—Hy tiers cannot downgrade once purchased.
Best for
High-frequency Hy3 agent developers and long-document analysis users
Usage
¥468/mo with 650M tokens—top Hy tier, below General Max (¥599) but Hy3 preview only in the pool.
Models
For full-stack AI generation, multi-agent collaboration, and CI automation with Hy3 as the core reasoning engine.
Highlights
Hy3 approaches frontier levels on benchmarks like SWE-bench Verified—cross-file refactors and long-doc analysis.
Enterprise Lite/Pro tiers below are a separate product line from personal Hy Max, with different URLs and key systems.
Enterprise Lite/Pro tiers below are a separate product line from personal Hy Max, with different URLs and key systems.
Best for
Heavy Hy3 agent users and professionals focused on Tencent in-house models
Usage
Monthly prepaid token subscription for teams—¥2/M tokens list; console picks pool size (min 50M ≈ ¥100/mo, max 5B/mo).
Models
Auto routing only (model id: auto)—no image/video multimodal yet; for teams needing lightweight unified routing.
Highlights
Multi-key: 1 key per 50M token quota—model access, total cap, and TPM per key; Base URL tokenhub.tencentmaas.com/plan/v3.
No downgrade or refund—keys expire if not renewed, tokens do not roll over; contact sales for larger quotas.
No downgrade or refund—keys expire if not renewed, tokens do not roll over; contact sales for larger quotas.
Best for
Teams needing Auto routing and multi-key allocation at moderate volume
企业专业
Team credit poolPrice
¥1,000/mo
100 credits = ¥1 · budget ¥1,000–¥20,000/mo (¥1,000 steps)
企业专业
Team credit poolPrice
¥1,000/mo
100 credits = ¥1 · budget ¥1,000–¥20,000/mo (¥1,000 steps)
Usage
Enterprise credit plan: 100 credits = ¥1, console budget ¥1,000–¥20,000 (¥1,000 steps), 1–12 month prepaid terms.
Models
13 models: Auto, GLM-5/5.1/Turbo, Kimi-K2.5/K2.6, MiniMax-M2.5/M2.7/M3, DeepSeek-V4-Flash/Pro and OEM ids—deducted by per-model credit rates.
Highlights
Independent credit pool per plan—multiple keys (1 per 10k credits), exclusive quota, total caps, and TPM for team governance.
Credits are not currency and non-transferable—no downgrade/refund; 2M credits/mo cap per plan, contact sales for more.
Credits are not currency and non-transferable—no downgrade/refund; 2M credits/mo cap per plan, contact sales for more.
Best for
Enterprises and R&D teams needing multi-model access, multi-key, and quota governance
Quota
Product type
Token Plan
Official product line
Token Plan
5-hour limit
—
Weekly limit
—
Monthly limit
35M–650M tokens/mo
Notes
- Personal list: General Lite ¥39–Max ¥599 (35M–650M tokens/mo); Hy Lite ¥28–Max ¥468, Hy3 only. One General + one Hy per account, one shared key.
- Enterprise Lite (131173): ¥2/M tokens list, 50M–5B token pool/mo (50M steps), ~¥100/mo minimum; Auto only (id: auto); 1 key per 50M tokens.
- Enterprise Pro (130659): 100 credits = ¥1; monthly budget ¥1,000–¥20,000 (¥1,000 steps), minimum 100k credits per purchase; 1–12 month terms; 1 key per 10k credits, 2M credits/mo cap per plan.
- General Token Plan sunsets Hunyuan-T1, Hunyuan-TurboS, and Tencent HY 2.0 series on 2026-06-22; kimi-k2.5 may throttle at peak—prefer other models when possible.
- Personal and enterprise: no refund/downgrade; unused quota does not roll over. Enterprise allows multiple plan packages per account—separate product line from personal.
Supported coding tools
OpenClawClaude CodeOpenCodeClineCursorKilo CodeCodex CLICodeBuddy
Pricing and model data sourced from official vendor websites
FAQ
General11 条