GLM-5.3 vs Hy3

Specs, VRAM requirements and download trends — updated 23 October 2026.

zai-org · A · Open weights
Params
753.3 B
Context
1,048,576
30d
1.9 M
tencent · B · Open weights
Params
298.8 B
Context
262,144
30d
798 K
apache-2.0 · commercial OK

Specification comparison

Differences are highlighted; identical values are muted.

Specification GLM-5.3 Hy3
Deployment Open weights · also via API Open weights · also via API
Parameters 753.3 B 298.8 B
Architecture GlmMoeDsaForCausalLM HYV3ForCausalLM
Context length 1,048,576 262,144
Licence other apache-2.0
Commercial use Unknown Allowed
Languages 2 —
Downloads 30d 1,937,342 797,888
Downloads all time 3.0 M 842 K
Quantizations on the Hub SAFETENSORS SAFETENSORS
Gated No No
First seen on the Hub 2026-08-25 2026-07-02
API input price / 1M tokens $0.079 $0.13
API output price / 1M tokens $1.40 $0.53
Est. cost per 1M input + 1M output tokens $1.48 $0.66
Cached input / 1M tokens $0.078 $0.033
AA Intelligence Index 44.8 —
AA Coding Index 74.8 —
AA Agentic Index 53.1 —
Max output tokens 943,718 128,000
API providers 37 5
Tool calling Yes Yes
Reasoning mode Yes Yes

API cost

Estimated cost of 1M input + 1M output tokens via OpenRouter

GLM-5.3 $1.48
Hy3 $0.66

Hy3 is 2.2× cheaper at list price.

Download trend

Daily snapshots, last 49 days (5 Sep – 23 Oct)

1.9 M188 K
GLM-5.3 Hy3

VRAM side by side

On RTX 4090 · 24 GB · 8K context unless noted

Hy3 · fp16 @ 256K ctx 2,092.0 GB / 24 GB
❌ Won’t fit
GLM-5.3 · fp16 @ 1024K ctx 15,295.6 GB / 24 GB
❌ Won’t fit

Adjacent comparisons