Skip to main content

On this page

DiscountsOther viewsColumn reference

Vendors

OpenAIAnthropicGooglexAIDeepSeekAlibabaByteDanceMoonshotZhipuMiniMaxAZ8BAAIBlack Forest LabsHappyHorseMicrosoft
Tiered pricing

Model Pricing Directory

Live pricing for all available models on APIYI: input/output/cache prices, endpoints, and groups. Auto-updated.

Data updated at 2026-10-07 13:51 (UTC+8) from the live pricing API. All prices in USD. Top-ups settle at a fixed 1:7 exchange rate, stackable with recharge bonuses — see Pricing.

Models online

323

Vendors

15

Usage-based

279

Per-call / per-second

44

Discounts

Prices in the tables below are default list prices — your effective cost can be lower, since recharge promotions and group discounts stack:
  • Recharge promotions: with recharge bonuses your effective cost drops to 1/1.1 of list price (about 91%), up to 1/1.2 (about 83%) during campaigns — see Recharge promotions.
  • Group discounts: some billing groups carry extra discounts (e.g. 0.95×, i.e. 95% of list) — see Group overview.
The console reflects the actual charge in real time.

Other views

Besides this page, model pricing is also available at:

Column reference

  • Type: the capability class of the model. Within each vendor table, rows are grouped by type and then ordered newest version first; dated snapshots and preview / thinking variants follow their base model.
  • Input / Output / Cache read: prices for usage-based models in USD per 1M tokens ($/1M). Cache read is the input price on prompt-cache hits; ”—” means no cache ratio is configured for the model.
  • Per-call: billed per request (image/video models), price in USD per call.
  • Per-second: billed by the length of the generated video (e.g. oxygen-1.0), price in USD per second of the requested duration.
  • Endpoints: chat = /v1/chat/completions; responses = /v1/responses; messages = /v1/messages; gemini = native /v1beta API; images = /v1/images/generations; embeddings = /v1/embeddings; videos = /v1/videos (submit and poll video tasks); hailuo = /hailuo/v2/video_generation (native MiniMax video API). A highlighted endpoint marked ★ is the recommended way to call the model.
  • Groups: billing groups where the model is available. Some groups carry discounts, stackable with recharge bonuses; the console is authoritative.
  • †: tiered pricing — the rate depends on per-request token volume, see “Tiered pricing” at the end of the page.
  • Linked model names: the model has its own detail page with context window, modalities, endpoint matrix and related docs.
  • This page is a complete single-page list — besides the search box above, desktop users can also use Ctrl+F (Cmd+F on macOS) to find a model name.
Jump to vendor: OpenAI · Anthropic · Google · xAI · DeepSeek · Alibaba · ByteDance · Moonshot · Zhipu · MiniMax · AZ8 · BAAI · Black Forest Labs · HappyHorse · Microsoft

OpenAI

For OpenAI GPT-5 and later, the native Responses API (/v1/responses) is recommended: reasoning and tool calling can be enabled together, built-in tools such as web search are available, and multi-turn prompt caching hits more reliably. Chat Completions still works; see the Responses API migration guide.

Anthropic

Google

xAI

DeepSeek

Alibaba

ByteDance

Moonshot

Zhipu

MiniMax

AZ8

BAAI

Black Forest Labs

HappyHorse

Microsoft

Tiered pricing

The following models are priced by per-request token volume (output price = tier input price × output multiplier):
  • gpt-6.1-sol: 0 – 272,000 tokens, input $2/1M; above 272,000 tokens, input $4/1M
  • gpt-6-astra: 0 – 272,000 tokens, input $10/1M; 272,001 – 1,050,000 tokens, input $20/1M
  • gpt-6-luna: 0 – 272,000 tokens, input $0.1/1M; above 272,000 tokens, input $0.2/1M
  • gpt-6-sol: 0 – 272,000 tokens, input $2/1M; above 272,000 tokens, input $4/1M
  • gpt-5.6-luna: 0 – 272,000 tokens, input $0.2/1M; above 272,000 tokens, input $0.4/1M
  • gpt-5.6-sol: 0 – 272,000 tokens, input $4/1M; above 272,000 tokens, input $8/1M
  • gpt-5.6-terra: 0 – 272,000 tokens, input $2/1M; above 272,000 tokens, input $4/1M
  • gpt-5.5-pro: 0 – 278,528 tokens, input $30/1M; above 278,528 tokens, input $60/1M
  • gpt-5.5: 0 – 278,528 tokens, input $5/1M; above 278,528 tokens, input $10/1M
  • gpt-5.4-pro: 0 – 278,528 tokens, input $30/1M; above 278,528 tokens, input $60/1M
  • gpt-5.4: 0 – 278,528 tokens, input $2.5/1M; above 278,528 tokens, input $5/1M
  • gemini-3.1-pro-preview: 0 – 200,000 tokens, input $1.8/1M; above 200,000 tokens, input $3.6/1M (includes a standing discount — 90% of the nominal list tiers)
  • grok-4.7: 0 – 200,000 tokens, input $2/1M; 200,001 – 500,000 tokens, input $4/1M
  • grok-4.6: 0 – 204,800 tokens, input $2/1M; 204,801 – 512,000 tokens, input $4/1M
  • grok-4.5: 0 – 204,800 tokens, input $2/1M; above 204,800 tokens, input $4/1M
  • grok-4.3: 0 – 204,800 tokens, input $1.25/1M; above 204,800 tokens, input $2.5/1M
  • qwen3.7-plus: 0 – 256,000 tokens, input $0.6/1M; 256,001 – 1,000,000 tokens, input $1.8/1M (includes a standing discount — 30% of the nominal list tiers)
  • qwen3.6-max-preview: 0 – 131,072 tokens, input $1.28/1M; 131,073 – 262,144 tokens, input $2.12/1M
  • qwen3.6-plus: 0 – 262,144 tokens, input $0.3/1M; 262,145 – 1,024,000 tokens, input $1.2/1M
  • qwen3.6-flash: 0 – 262,144 tokens, input $0.17/1M; 262,145 – 1,024,000 tokens, input $0.68/1M
  • qwen3.5-plus: 0 – 131,072 tokens, input $0.114/1M; 131,073 – 262,144 tokens, input $0.28/1M; above 262,144 tokens, input $0.56/1M
  • qwen3.5-plus-2026-02-15: 0 – 131,072 tokens, input $0.114/1M; 131,073 – 262,144 tokens, input $0.28/1M; above 262,144 tokens, input $0.56/1M
  • qwen3.5-flash: 0 – 131,072 tokens, input $0.03/1M; 131,073 – 262,144 tokens, input $0.122143/1M; above 262,144 tokens, input $0.182143/1M
  • qwen3.5-flash-2026-02-23: 0 – 131,072 tokens, input $0.03/1M; 131,073 – 262,144 tokens, input $0.122143/1M; above 262,144 tokens, input $0.182143/1M
  • qwen3-max: 0 – 32,000 tokens, input $1.2/1M; 32,001 – 128,000 tokens, input $1.92/1M; 128,001 – 256,000 tokens, input $3.36/1M (includes a standing discount — 48% of the nominal list tiers)
  • qwen3-max-preview: 0 – 32,000 tokens, input $1.2/1M; 32,001 – 128,000 tokens, input $2/1M; 128,001 – 256,000 tokens, input $3/1M (includes a standing discount — 20% of the nominal list tiers)
  • qwen3-max-2025-09-23: 0 – 32,000 tokens, input $1.2/1M; 32,001 – 128,000 tokens, input $2/1M; 128,001 – 256,000 tokens, input $3/1M (includes a standing discount — 20% of the nominal list tiers)
  • qwen3-coder-plus: 0 – 32,000 tokens, input $5/1M; 32,001 – 128,000 tokens, input $7.5/1M; 128,001 – 256,000 tokens, input $12.5/1M; 256,001 – 1,000,000 tokens, input $25/1M
  • qwen3-coder-plus-2025-09-23: 0 – 32,000 tokens, input $2/1M; 32,001 – 128,000 tokens, input $3/1M; 128,001 – 256,000 tokens, input $5/1M; 256,001 – 1,000,000 tokens, input $10/1M (includes a standing discount — 50% of the nominal list tiers)
  • qwen3-coder-plus-2025-07-22: 0 – 32,000 tokens, input $2/1M; 32,001 – 128,000 tokens, input $3/1M; 128,001 – 256,000 tokens, input $5/1M; 256,001 – 1,000,000 tokens, input $10/1M (includes a standing discount — 50% of the nominal list tiers)
  • qwen3-coder-480b-a35b-instruct: 0 – 32,768 tokens, input $3/1M; 32,769 – 131,072 tokens, input $5.4/1M; 131,073 – 204,800 tokens, input $9/1M
  • qwen3-coder-flash: 0 – 32,000 tokens, input $0.5/1M; 32,001 – 128,000 tokens, input $0.75/1M; 128,001 – 256,000 tokens, input $1.25/1M; 256,001 – 1,000,000 tokens, input $2.5/1M (includes a standing discount — 50% of the nominal list tiers)
  • qwen3-vl-plus: 0 – 32,000 tokens, input $0.3/1M; 32,001 – 128,000 tokens, input $0.45/1M; 128,001 – 256,000 tokens, input $0.9/1M (includes a standing discount — 30% of the nominal list tiers)
  • qwen3-vl-plus-2025-09-23: 0 – 32,000 tokens, input $0.3/1M; 32,001 – 128,000 tokens, input $0.45/1M; 128,001 – 256,000 tokens, input $0.9/1M (includes a standing discount — 30% of the nominal list tiers)
  • qwen-plus: 0 – 128,000 tokens, input $0.4/1M; 128,001 – 256,000 tokens, input $1.2/1M; 256,001 – 1,000,000 tokens, input $2.4/1M (includes a standing discount — 50% of the nominal list tiers)
  • qwen-plus-latest: 0 – 128,000 tokens, input $0.4/1M; 128,001 – 256,000 tokens, input $1.2/1M; 256,001 – 1,000,000 tokens, input $2.4/1M (includes a standing discount — 50% of the nominal list tiers)
  • qwen-plus-2025-09-11: 0 – 128,000 tokens, input $0.4/1M; 128,001 – 256,000 tokens, input $1.2/1M; 256,001 – 1,000,000 tokens, input $2.4/1M (includes a standing discount — 50% of the nominal list tiers)
  • seed-2-0-pro-260328: 0 – 131,072 tokens, input $0.5/1M; 131,073 – 262,144 tokens, input $1/1M
  • seed-2-0-lite-260228: 0 – 131,072 tokens, input $0.25/1M; 131,073 – 262,144 tokens, input $0.5/1M
  • seed-2-0-mini-260215: 0 – 131,072 tokens, input $0.1/1M; 131,073 – 262,144 tokens, input $0.2/1M
  • glm-5.1: 0 – 32,768 tokens, input $0.84/1M; above 32,768 tokens, input $1.14/1M
  • glm-5: 0 – 32,000 tokens, input $0.56/1M; above 32,000 tokens, input $0.86/1M
  • MiniMax-M3: 0 – 524,288 tokens, input $0.3/1M; above 524,288 tokens, input $0.6/1M (includes a standing discount — 50% of the nominal list tiers)
Prices are subject to change; the console reflects live rates. See the API Capabilities section for per-model usage guides.