GLM-5.3 vs Kimi K3: API pricing and features
GLM-5.3 vs Kimi K3 on BigModel.io: price, context, inputs and Claude Code support. GLM-5.3 $0.20 / $0.64, Kimi K3 $2.70 / $13.50 per 1M input / output tokens.
In short
- GLM-5.3 is cheaper: a typical coding-agent request (20K in, 2K out) costs about $0.0053, against $0.081 on Kimi K3, which costs about 15.3 times as much.
- Both work in Claude Code.
- Only Kimi K3 accepts image input.
Prices are what BigModel.io charges today, in US dollars. Quality depends on your task: try both on the same prompt before you switch.
Side by side
| GLM-5.3 | Kimi K3 | |
|---|---|---|
| Model ID | glm-5.3 | kimi-k3 |
| Developer | GLM | Kimi |
| Input / 1M tokens | $0.20 | $2.70 |
| Output / 1M tokens | $0.64 | $13.50 |
| Cached input / 1M | $0.04 | $0.30 |
| Below official price | 85% | 10% |
| Agent request (20K in, 2K out) | $0.0053 | $0.081 |
| Context window | 1M | — |
| Inputs | Text | Text, Image, Video |
| Claude Code (/v1/messages) | Yes | Yes |
| Chat Completions | Yes | Yes |
Switch between them
Both use the same key and base URL. Change only the model field:
client.chat.completions.create(model="glm-5.3", ...) # or "kimi-k3"To fall back automatically when one is unavailable, send both in models: "models": ["glm-5.3", "kimi-k3"] Fallback →