GLM 4.5 Air
GLM 4.5 AirRankingin thinking mode?
Alternatives?
CheaperClaude Haiku 5.5max×3.9-53%$0.21Remove sampling parameters+5
BetterGPT-6 Lunalow+94%-53%$0.005Function calling only with none · Chat Completions+3
OpenQwen3.8 Flashthinking×3.6-46%$0.37Memory (4-bit): 66 GB → 108 GB+6
CheaperMiniMax M3.1 Flashthinking×3.9-38%$0.22Complete model response must be appendedCheaperGLM 5.3 FlashmaxOpen×3.8-44%$0.28Memory (4-bit): 66 GB → 193 GB+5Claude Haiku 5.5max×3.9-53%$0.21Remove sampling parameters+5MiniMax M3.1 Flashthinking×3.9-38%$0.22Complete model response must be appendedGLM 5.3 FlashmaxOpen×3.8-44%$0.28Memory (4-bit): 66 GB → 193 GB+5Qwen3.8 FlashthinkingOpen×3.6-46%$0.37Memory (4-bit): 66 GB → 108 GB+6GPT-6 Lunamax×3.4-53%$0.068Function calling only with none · Chat Completions+3GPT-6 Lunalow+94%-53%$0.005Function calling only with none · Chat Completions+3GPT-5.6 Lunalow+89%+6%$0.01
MiMo v2.5thinkingOpen+97%-59%$0.019Memory (4-bit): 66 GB → 186 GBClaude Haiku 5.5low×2.7-53%$0.024Remove sampling parameters+5GPT-5 Minihigh+51%+62%$0.053Qwen3.8 FlashthinkingOpen×3.6-46%$0.37Memory (4-bit): 66 GB → 108 GB+6Qwen3.8 27BxhighOpen×3+125%$1.01max_tokens range · QwenCloud+4
Nemotron 3.5 Super VLOpen×2.4-59%$0.19Memory (4-bit): 66 GB → 75 GBTInkling SmallthinkingOpen×2.3+50%$0.089Memory (4-bit): 66 GB → 160 GBIK2 Horizon MoVA 36B A4BthinkingOpen×2.3+312%$1 / $4GLM 5.3 FlashmaxOpen×3.8—$0.28Memory (4-bit): 66 GB → 193 GB+5GLM 5.3lowOpen×3.1—$0.85Memory (4-bit): 66 GB → 452 GB+4GLM 5thinkingOpen×2.5—$1 / $3.20Memory (4-bit): 66 GB → 452 GBGLM 5 Turbothinking×2.4—$1.20 / $4GLM 5.1thinkingOpen×2.3—$1.02Memory (4-bit): 66 GB → 452 GBGLM 5V Turbothinking×2.1—$1.20 / $4GLM 5.2non-thinkingOpen×2—$1.40 / $4.40Memory (4-bit): 66 GB → 452 GBGLM 4.7thinkingOpen×2—$0.60 / $2.20Memory (4-bit): 66 GB → 215 GBAll Z.ai models →Run it yourself
estimateBenchmarks
all categoriesFAQ
How much does GLM 4.5 Air cost?›
$0.20 per 1M input tokens and $1.10 per 1M output tokens; cached input $0.03 per 1M; cheapest via Novita at $0.13 / $0.85.
Is GLM 4.5 Air free to use?›
There is no free API; the cheapest paid price is $0.13 per 1M input tokens. The weights are open, so you can run it yourself.
How good is GLM 4.5 Air, and what is it best at?›
#221 of 334 in Chat & Knowledge, #90 of 239 in Coding, #116 of 242 in Agents & Tool Use, #114 of 195 in Writing & Language.
Is GLM 4.5 Air better than Qwen3 Next 80B A3B Thinking?›
On AA Intelligence Index 4.3.2, GLM 4.5 Air scores 11.1 vs 11.2 — about the same.
How can I use GLM 4.5 Air?›
Through the API of DeepInfra, Novita, SiliconFlow.
What is the context window of GLM 4.5 Air?›
131K tokens, with up to 98K output tokens.
When was GLM 4.5 Air released? Is there a newer version?›
Released 2025-07-28. It has been superseded by GLM 4.7 Flash.
Is GLM 4.5 Air open source?›
Yes, the weights are open (MIT licence).
Can I run GLM 4.5 Air locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can GLM 4.5 Air read images?›
No, it accepts text only.
What is GLM 4.5 Air?›
GLM 4.5 Air is an open-weights language model from Z.ai, released Jul 2025.
What are the best alternatives to GLM 4.5 Air?›
Claude Haiku 5.5 (cheaper), GPT-6 Luna (better), Qwen3.8 Flash (open weights).