Skip to content

Hy4 preview vs GLM-5.3

Hy4 preview is from Tencent, released August 28, 2026. GLM-5.3 is from Zhipu AI, released August 18, 2026.

  • Input per 1M$0.834 vs $1.40. Hy4 preview costs 40% less per million input tokens.
  • Output per 1M$2.501 vs $4.40. Hy4 preview costs 43% less per million output tokens.
  • Context window1M vs 1M. Both hold 1M tokens of context.

Prices are the vendors’ published API rates; Elo is LMArena’s text board. Each figure is dated on its model page.

Interactive Comparison

Loading chart...
$0.834 / $2.501 per million tokens
$1.40 / $4.40 per million tokens
Identity
DeveloperTencentZhipu AI
ReleasedAug 2026Aug 2026
StatusPreviewGA
LicenceApache 2.0 (open weights)GLM-5.3 License (custom: MIT-style terms plus a security-review clause for MaaS providers above $10B revenue); weights on Hugging Face, zai-org/GLM-5.3
Self-hostableYes — tencent/Hy4-preview on Hugging FaceYes - weights published on Hugging Face
Cost
Blended $/1M tokensinput × 0.75 + output × 0.25$1.251 / 1M tokens$2.150 / 1M tokens
Input price$0.834 / 1M tokens$1.400 / 1M tokens
Output price$2.501 / 1M tokens$4.400 / 1M tokens
Cached input$0.042 / 1M tokens$0.26 / 1M tokens
Batch discount—Not offered; GLM Coding Plan applies a 50% off-peak points reduction
Free tier—Not offered
Capacity
Context window1M tokens1M tokens
Max output64K tokens128k tokens
Long-context surcharge—Not published
Capability
Vision in—No — text-only input
Audio in—No
Function callingYesYes
Structured outputYesYes
Extended reasoningYes — deep reasoningYes — always on, three effort levels (low, high, max)
Web search—Not documented
Code execution—Not documented
Access
APIYes — Tencent Cloud TokenHub (hy4-preview)Yes
Chat app—Yes
Cloud marketplaces—Not offered
Fine-tuning—Not offered
Measured quality
LMArena Elo—1481 (checked Oct 2026)
LMArena rank—Rank 26
GPQA Diamond—91
Elo per dollarLMArena Elo ÷ blended $/1M tokens—689

Add a third model to this comparison