GLM 4.5V
GLM 4.5VRankingin non-thinking mode?
Alternatives?
CheaperKimi K2 ThinkingthinkingOpen×6.2-5%$0.47 / $2Memory (4-bit): 65 GB → 616 GB+1
BetterMinistral 3 3B 2512Open+43%-89%$0.008
OpenGPT OSS 120Bhigh×6.1-93%$0.11Memory (4-bit): 65 GB → 70 GB+1
CheaperNova 2.0 Litehigh×6.2-6%$0.30 / $2.50No strict JSON
CheaperNemotron 3 Nano 30B A3BthinkingOpen×5.9-88%$0.017No image inputKimi K2 ThinkingthinkingOpen×6.2-5%$0.47 / $2Memory (4-bit): 65 GB → 616 GB+1Nova 2.0 Litehigh×6.2-6%$0.30 / $2.50No strict JSONGPT OSS 120BhighOpen×6.1-93%$0.11Memory (4-bit): 65 GB → 70 GB+1Nemotron 3 Nano 30B A3BthinkingOpen×5.9-88%$0.017No image input
Qwen3 235B A22B Thinking 2507thinkingOpen×5.9-17%$0.08Memory (4-bit): 65 GB → 141 GB+2Ministral 3 3B 2512Open+43%-89%$0.008GPT OSS 20BhighOpen×5.8-96%$0.011No image inputMinistral 3 8B 2512Open×2.1-83%$0.011Nemotron 3 Nano 30B A3BthinkingOpen×5.9-88%$0.017No image inputMinistral 3 14B 2512non-thinkingOpen+96%-78%$0.019GPT OSS 120BhighOpen×6.1-93%$0.11Memory (4-bit): 65 GB → 70 GB+1Nemotron 3 Nano 30B A3BthinkingOpen×5.9-88%$0.017No image inputQwen3 235B A22B Thinking 2507thinkingOpen×5.9-17%$0.08Memory (4-bit): 65 GB → 141 GB+2
K EXAONEthinkingOpen×5.9——Memory (4-bit): 65 GB → 142 GB+1GPT OSS 20BhighOpen×5.8-96%$0.011No image inputGLM 4.7thinkingOpen×6.2—$0.60 / $2.20Memory (4-bit): 65 GB → 215 GB+1GLM 4.5 AirthinkingOpen×5.3—$0.20 / $1.10Memory (4-bit): 65 GB → 66 GB+2GLM 4.5thinkingOpen×4.8—$0.60 / $2.20Memory (4-bit): 65 GB → 215 GB+1GLM-4.6non-thinkingOpen×2.9—$0.60 / $2.20Memory (4-bit): not verified+1GLM 4.6Vnon-thinkingOpen+72%—$0.30 / $0.90GLM 5.3 FlashXOpenn/a—$0.37 / $1.25Memory (4-bit): not verifiedGLM 5.3 FlashmaxOpenn/a—$0.28Memory (4-bit): 65 GB → 193 GB+5GLM 5.3lowOpenn/a—$0.85Memory (4-bit): 65 GB → 452 GB+5All Z.ai models →Run it yourself
estimateHistory
10 versionsBenchmarks
all categoriesFAQ
How much does GLM 4.5V cost?›
$0.60 per 1M input tokens and $1.80 per 1M output tokens; cached input $0.11 per 1M.
Is GLM 4.5V free to use?›
There is no free API; the cheapest paid price is $0.60 per 1M input tokens. The weights are open, so you can run it yourself.
How good is GLM 4.5V, and what is it best at?›
#327 of 336 in Chat & Knowledge, #142 of 239 in Coding, #193 of 242 in Agents & Tool Use, #128 of 197 in Writing & Language, #118 of 125 in Vision & Documents.
Is GLM 4.5V better than Qwen3 32B?›
On AIME 2025 · AA archived, GLM 4.5V scores 73 vs 73 — about the same.
How can I use GLM 4.5V?›
Through the API of Novita.
What is the context window of GLM 4.5V?›
66K tokens, with up to 16K output tokens.
When was GLM 4.5V released? Is there a newer version?›
Released 2025-08-11. It has been superseded by GLM 4.6V.
Is GLM 4.5V open source?›
Yes, the weights are open (MIT licence).
Can I run GLM 4.5V locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can GLM 4.5V read images?›
It accepts text and images as input.
What is GLM 4.5V?›
GLM 4.5V is an open-weights language model from Z.ai, released Aug 2025.
What are the best alternatives to GLM 4.5V?›
Kimi K2 Thinking (cheaper), Ministral 3 3B 2512 (better), GPT OSS 120B (open weights).