# Long-context Chinese models: what 1M tokens actually costs (markdown mirror)

Source: https://aiapi-pro.com/blog/long-context-chinese-models-cost-2026

"Full pass" = input price per 1M tokens x window size: the cost of one request that fills the window.
Prices are provider list price with a 1.0 multiplier (0% platform markup). Feed: https://aiapi-pro.com/china-ai-leaderboard.json

## Cheapest model per tier

| Window | Cheapest model | Input / 1M | Cost per full pass |
|---|---|---|---|
| 4M | MiniMax Text-01 | $0.140 | $0.56 |
| 1M | GLM-5.3-Flash | $0.080 | $0.08 |
| 512K | MiniMax M3 | $0.300 | $0.15 |
| 256K | Doubao-Seed-2.0-Lite | $0.075 | $0.02 |
| 8K | Hunyuan MT2 Lite | $0.043 | $0.0003 |

## All long-window models

| Window | Model | Provider | Input / 1M | Full pass |
|---|---|---|---|---|
| 4M | MiniMax Text-01 | MiniMax | $0.140 | $0.56 |
| 1M | GLM-5.3-Flash | Zhipu AI | $0.080 | $0.08 |
| 1M | MiMo V2.5 Pro | Xiaomi | $0.427 | $0.43 |
| 1M | Hunyuan Hy4 Preview | Tencent | $0.834 | $0.83 |
| 1M | GLM-5.3 | Zhipu AI | $1.250 | $1.25 |
| 1M | Kimi K3 | Moonshot | $2.900 | $2.90 |
| 512K | MiniMax M3 | MiniMax | $0.300 | $0.15 |
| 256K | Doubao-Seed-2.0-Lite | ByteDance | $0.075 | $0.02 |
| 256K | Hunyuan Hy3 Preview | Tencent | $0.127 | $0.03 |
| 256K | Qwen3 Coder Flash | Alibaba Cloud | $0.150 | $0.04 |
| 256K | Doubao Seed 2.1 Turbo | ByteDance | $0.150 | $0.04 |
| 256K | Hunyuan Hy3 | Tencent | $0.180 | $0.05 |
| 256K | Qwen3 VL Plus | Alibaba Cloud | $0.210 | $0.05 |
| 256K | MiniMax M2.5 | MiniMax | $0.270 | $0.07 |
| 256K | MiniMax M2.7 | MiniMax | $0.300 | $0.08 |
| 256K | Doubao-Seed-2.0-Pro | ByteDance | $0.400 | $0.10 |
| 256K | Kimi K2.5 | Moonshot | $0.600 | $0.15 |
| 256K | Kimi K2.7 Code | Moonshot | $0.670 | $0.17 |
| 256K | Kimi K2.8 Preview | Moonshot | $0.930 | $0.24 |
| 256K | Kimi K2.6 | Moonshot | $0.950 | $0.24 |
| 256K | Kimi K2.7 Code HighSpeed | Moonshot | $2.073 | $0.53 |
| 202K | GLM-4.7-Flash | Zhipu AI | FREE | $0 |
| 128K | DeepSeek V4 Flash | DeepSeek | $0.080 | $0.01 |
| 128K | DeepSeek V4 Pro | DeepSeek | $0.570 | $0.07 |

## Three findings

1. **Context size and price are independent.** Doubao-Seed-2.0-Lite: 256K at $0.075/1M. Kimi K2.7 Code HighSpeed: 256K at $2.073/1M. 28x for the same capacity.
2. **The cheapest 1M window is not a flagship.** GLM-5.3-Flash at $0.08/1M = $0.08 per full pass; Kimi K3 at $2.90.
3. **4M exists and is not the priciest.** MiniMax Text-01: $0.14/1M, $0.56 per full 4M pass - less than one full pass on Kimi K3.

## What a full pass buys

- ~128K tokens: a 300-page book or a mid-sized repo's relevant files
- ~1M tokens: a large codebase or a full documentation set
- ~4M tokens: multiple books or a small product's source history

Most long-context requests fill 10-20% of the window, cutting these figures 5-10x.

## FAQ

**What is the cheapest 1M-context model?**
GLM-5.3-Flash at $0.08 per 1M input tokens, so one full-context pass costs about $0.08. Kimi K3 also offers 1M context but at $2.90 per 1M input, roughly 36x more for the same volume.

**Which model has the largest context window?**
MiniMax Text-01 lists a 4M-token context window, the largest in the catalog. A full 4M-token pass costs about $0.56 at its $0.14 per 1M input price.

**Does a bigger context window always cost more?**
No. Context size and price are independent. Doubao-Seed-2.0-Lite is the cheapest 256K model at $0.075 per 1M input, while Kimi K2.7 Code HighSpeed costs $2.073 per 1M input at the same 256K window.

**Are these prices per request or per token?**
Per token. You pay for the tokens you send and receive, not for the size of the window. A 10K-token request on a 1M-context model costs 1% of a full-context pass.