Discounted AI Models Up to 2 折
Direct official access · curated models · discounted rates
Series currently on discount: gpt · gemini · seedance · glm · doubao
Current discounts
Sorted by discount, largest first, synced with official vendor pricing. Struck-through prices are official list prices; orange is the price you pay. Promotions follow vendor changes and may end at any time.
| Model | Discount | Input | Output | Cached input | Capabilities |
|---|---|---|---|---|---|
openai/gpt-5.6-luna | 2 折 | $0.2/M$1/M | $1.2/M$6/M | $0.02/M$0.1/M | 文本推理视觉 |
google/gemini-3.7-flash | 5 折 | $0.75/M$1.5/M | $3.75/M$7.5/M | $0.075/M$0.15/M | 文本推理视觉 |
google/gemini-3.6-flash | 5 折 | $0.75/M$1.5/M | $3.75/M$7.5/M | $0.075/M$0.15/M | 文本推理视觉 |
OpenAI: GPT-5.6 SolLMArena #10 openai/gpt-5.6-sol | 5 折 | $2.5/M$5/M | $15/M$30/M | $0.25/M$0.5/M | 文本推理视觉 |
bytedance/seedance-2.0-mini | 5 折 | — | $0.02/s 起$0.04/s 起 | — | 生视频 |
Z.ai: GLM-5.2LMArena #35 z-ai/glm-5.2 | 7 折 | $0.98/M$1.4/M | $3.08/M$4.4/M | $0.182/M$0.26/M | 文本推理工具调用 |
bytedance/seedance-2.0-fast | 7 折 | — | $0.042/s 起$0.06/s 起 | — | 生视频 |
bytedance/seedance-2.5 | 8 折 | — | $0.568/s · 1080p$0.71/s · 1080p | — | 生视频 |
openai/gpt-5.6-terra | 8 折 | $2/M$2.5/M | $12/M$15/M | $0.2/M$0.25/M | 文本推理视觉 |
volcengine/doubao-seed-2.1-pro | 8 折 | $0.7072/M$0.884/M | $3.536/M$4.42/M | $0.1416/M$0.177/M | 文本推理视觉 |
volcengine/doubao-seed-2.1-turbo | 8 折 | $0.3536/M$0.442/M | $1.7696/M$2.212/M | $0.068/M$0.085/M | 文本推理视觉 |
bytedance/seedance-2.0 | 9 折 | — | $0.063/s 起$0.07/s 起 | — | 生视频 |
z-ai/glm-5.3 | 9 折 | $1.26/M$1.4/M | $3.96/M$4.4/M | $0.234/M$0.26/M | 文本推理工具调用 |
Rankings from LMArena (2026-07-12, CC BY 4.0) lmarena.ai
Browse by use case
The same discounted models, grouped by capability — tool calling, context window, modality.
Coding & agents
Tool calling and reasoning, compatible with Claude Code, Codex and Cline. Discounted rates apply to agent workloads as well.
Set up your coding tool →Long documents & RAG
500K+ context windows at the same discounted rates — whole-repo and long-document workloads.
Compare long-context models →High volume, low cost
The lowest current output prices among the discounted chat models.
Find the cheapest fit →Best value at list price
The lowest standing output prices per 1M tokens — standard pricing, no promotion.
Ready to build at discounted rates ?
3 minutes to integrate — discounts apply automatically.
Get API Key