GLM-5.3 Pricing, Benchmarks & Channels
GLM-5.3 is a 2026-08-14 model from Mistral AI featuring a 1M context window and extended reasoning capabilities. Lowest rate on OpenRouter ($0.084 in / $3.6 out per 1M tokens).
Input Price / 1M
$1.4
Output Price / 1M
$4.4
Cache Read / 1M
$0.14
Context Window
1M
Generation Speed
86 TPS
Access Grade
A · Global
F.A.R.E. Four-Dimension Score Breakdown
Scenario Capability Breakdown (0–100):
💻 Coding
93/100
✍️ Writing
94/100
📚 Long-Doc RAG
95/100
👁️ Vision OCR
Text-only
F.A.R.E. CAPABILITY RADAR
GLM-5.3 Multi-Platform Price Comparison (10 Verified Channels)
All prices are in USD per 1M tokens. Domestic developers can prioritize channels marked "Direct in CN" for compliant access and local billing.
| Platform / Gateway | Network Access | Input ($/1M) | Output ($/1M) | Cache Hit ($/1M) | Billing & Notes | Link |
|---|---|---|---|---|---|---|
| OpenRouter Lowest Price | Direct in CN | $0.084 | $3.6 | $0.079 | Direct CN access · RMB billing | Visit Official → |
| DeepInfra | Global Network | $0.9 | $4 | $0.2 | Global cloud billing | Visit Official → |
| Alibaba Cloud Bailian (CN) | Direct in CN | $1.1 | $3.851 | $0.275 | Direct CN access · RMB billing | Visit Official → |
| Zhipu BigModel (CN) | Direct in CN | $1.1952 | $4.1831 | $0.2988 | Direct CN access · RMB billing | Visit Official → |
| SiliconFlow (CN) | Direct in CN | $1.1995 | $4.1982 | $0.2999 | Direct CN access · RMB billing | Visit Official → |
| Mistral Official | Global Network | $1.4 | $4.4 | $0.14 | Global cloud billing | Visit Official → |
| 302.AI Gateway | Direct in CN | $1.4 | $4.4 | $0.26 | Direct CN access · RMB billing | Visit Official → |
| Together AI | Global Network | $1.4 | $4.4 | $0.26 | Global cloud billing | Visit Official → |
| Fireworks AI | Global Network | $1.4 | $4.4 | $0.26 | Global cloud billing | Visit Official → |
| AWS Bedrock | Global Network | $1.848 | $5.808 | $0.3432 | Global cloud billing | Visit Official → |
✅ Core Strengths & Ideal Workloads
- Top-tier empirical capability score (94 index)
- Open weights available for private self-hosting (~74GB VRAM at INT4)
⚠️ Limitations & Practical Caveats
- European frontier lab available via Mistral La Plateforme, AWS Bedrock, and OpenRouter.
- Extended thinking generates internal reasoning tokens (~1.8x output multiplier)