Global AI Token Comparison Hub
CN
Back to all plans

火山引擎方舟

Multi-model routing

Ark cross-tool Coding Plan (not an IDE)—Doubao Seed 2.0 plus GLM-5.2, MiniMax-M3, Kimi, and DeepSeek-V4 in one shared pool

Coding PlanCoding Plan
SubscriptionToken API
Official

Core models

AutoDoubao-Seed-2.0-CodeDoubao-Seed-2.0-ProDoubao-Seed-2.0-LiteDoubao-Seed-CodeMiniMax-M2.7MiniMax-M3Kimi-K2.6GLM-5.2DeepSeek-V4-FlashDeepSeek-V4-Pro
Auto

Auto routing via ark-code-latest—platform picks the best model mix for quality and speed.

Doubao-Seed-2.0-Code

doubao-seed-2.0-code—Doubao coding flagship with vision and strong frontend generation for agents and full-stack coding.

Doubao-Seed-2.0-Pro

doubao-seed-2.0-pro—Doubao 2.0 flagship for complex reasoning and long-chain agents.

Doubao-Seed-2.0-Lite

doubao-seed-2.0-lite—lightweight Doubao 2.0 for daily completion and medium-complexity tasks.

Doubao-Seed-Code

doubao-seed-code—prior Doubao code line focused on debugging and refactors; switch with 2.0-Code by task.

See the official site for more models

Additional core model names still appear above, with full details on the latest official page.

Open official site

Plan details

Lite

Medium
Price
¥40/mo
1,200/5h · 9,000/wk · 18,000/mo calls
Monthly
¥40/mo
Auto quarterly
¥40/mo
¥120/qtr
Official
Usage
Lite at ¥40/mo includes ~1,200/5h, 9,000/wk, and 18,000/mo request quota—for personal projects, learning, and medium-intensity daily coding.
Models
Supports 11 models plus Auto: Doubao Seed 2.0 Code/Pro/Lite, GLM-5.2, MiniMax-M3/M2.7, Kimi-K2.6, DeepSeek-V4-Flash/Pro, and more—switch Model Name live in tools.
Highlights
5-hour window rolls from first request; weekly resets Monday 00:00; monthly resets on subscription day 1 00:00—when exhausted, wait for refresh without consuming account balance.
Suited to first-time multi-model routing and comparing vendors in Cursor or Claude Code before deciding on Pro.
Best for
Medium-intensity individual developers exploring multi-model AI coding

Pro

PowerRecommended
Price
¥200/mo
6,000/5h · 45,000/wk · 90,000/mo calls
Monthly
¥200/mo
Auto quarterly
¥200/mo
¥600/qtr
Official
Usage
Pro at ¥200/mo is 5× Lite: ~6,000/5h, 45,000/wk, 90,000/mo—for developers who treat AI coding as a primary workbench.
Models
Same 11 models plus Auto; higher Pro TPM for peak stability—better for complex agent orchestration and multi-repo work.
Highlights
2.5× promo (~¥9.9/¥49.9 first two months) stacks with 9.5% referral discount—useful for sustained trial before long-term renewal.
Quota shared across tools; ark-code-latest lets you switch models in the console within 3–5 minutes without editing tool configs.
Best for
Heavy coders, complex agent workflows, and professionals using multi-model subscriptions as primary tools

Quota

Product type
Coding Plan
Official product line
Coding Plan
5-hour limit
1,200–6,000 calls/5h
Weekly limit
9,000–45,000 calls/wk
Monthly limit
18,000–90,000 calls/mo

Notes

  • List price Lite ¥40/mo (¥120/qtr), Pro ¥200/mo (¥600/qtr) per activity and pricing pages; Lite & Pro 2.5× promo (~¥9.9/¥49.9) for first two months on new buy/renew/upgrade, then list rate—limited slots.
  • Referrals: friend gets extra 9.5% off first order, referrer earns 5% voucher; Jun 10–30, 2026 extra 2.5× deduction discount on deepseek-v4-pro, kimi-k2.6, glm-5.2—see activity docs.
  • Quota is for coding tools only—no API scripts, backends, or non-interactive batch calls; enterprise workloads should use Ark pay-as-you-go API.
  • One Coding Plan subscription shares quota across all supported tools—usage in Claude Code deducts from the same pool as OpenClaw and others, no duplicate purchases needed.

Supported coding tools

TRAEClaude CodeCursorClineOpenCodeOpenClawRoo CodeKilo CodeCodex CLIHermes Agent

Pricing and model data sourced from official vendor websites

FAQ

General·9
General
9 条