GLM 4.7 in Cursor

[Somhairle_Marisol (Somhairle H. Marisol)]

Feature request for product/service

AI Models

Describe the request

GLM 4.7 seems like a practical and cost-effective model—please add it to Cursor’s officially supported model list.

Benchmark Performance. More detailed comparisons of GLM-4.7 with other models GPT-5, GPT-5.1-High, Claude Sonnet 4.5, Gemini 3.0 Pro, DeepSeek-V3.2, Kimi K2 Thinking, on 17 benchmarks (including 8 reasoning, 5 coding, and 3 agents benchmarks) can be seen in the below table.

Benchmark	GLM-4.7	GLM-4.6	Kimi K2 Thinking	DeepSeek-V3.2	Gemini 3.0 Pro	Claude Sonnet 4.5	GPT-5 High	GPT-5.1 High
Reasoning
MMLU-Pro	84.3	83.2	84.6	85.0	90.1	88.2	87.5	87.0
GPQA-Diamond	85.7	81.0	84.5	82.4	91.9	83.4	85.7	88.1
HLE	24.8	17.2	23.9	25.1	37.5	13.7	26.3	25.7
HLE (w/ Tools)	42.8	30.4	44.9	40.8	45.8	32.0	35.2	42.7
AIME 2025	95.7	93.9	94.5	93.1	95.0	87.0	94.6	94.0
HMMT Feb. 2025	97.1	89.2	89.4	92.5	97.5	79.2	88.3	96.3
HMMT Nov. 2025	93.5	87.7	89.2	90.2	93.3	81.7	89.2	-
IMOAnswerBench	82.0	73.5	78.6	78.3	83.3	65.8	76.0	-
LiveCodeBench-v6	84.9	82.8	83.1	83.3	90.7	64.0	87.0	87.0
Code Agent
SWE-bench Verified	73.8	68.0	73.4	73.1	76.2	77.2	74.9	76.3
SWE-bench Multilingual	66.7	53.8	61.1	70.2	-	68.0	55.3	-
Terminal Bench Hard	33.3	23.6	30.6	35.4	39.0	33.3	30.5	43.0
Terminal Bench 2.0	41.0	24.5	35.7	46.4	54.2	42.8	35.2	47.6
General Agent
BrowseComp	52.0	45.1	-	51.4	-	24.1	54.9	50.8
BrowseComp (w/ Context Manage)	67.5	57.5	60.2	67.6	59.2	-	-	-
BrowseComp-ZH	66.6	49.5	62.3	65.0	-	42.4	63.0	-
τ²-Bench	87.4	75.2	74.3	85.3	90.7	87.2	82.4	82.7

GLM-4.7: Advancing the Coding Capability

[CFalcon075]

Yes, we need this!!

[DonNguyen]

It seems like another attempt, but GLM (China) doesn’t appear to have been included in the cursor development process.

[user616 (361846570)]

Actually , GLM’s model didn’t work well in reality , I am a Chinese , I bought GLM 4.5 last month . The capabilities of this model have been exaggerated.

[Somhairle_Marisol (Somhairle H. Marisol)]

Thanks for sharing your real-world experience! GLM 4.5 seems like a pretty mediocre model—it’s cheap and just okay performance-wise. But after two versions, GLM 4.7 now costs ($0.40/M input tokens, $1.50/M output tokens), which isn’t really more expensive than 4.5 ($0.35/M input tokens, $1.55/M output tokens), and it performs way better. I’d say it’s worth using.

[user616 (361846570)]

Their model is based on the MOE architecture, which is why it is so cheap. What I mean is, their model may appear very impressive on those programming leaderboards, but in actual use, it has its flaws. In contrast, Claude Sonnet or GPT 5 are much more versatile.

[Shivam_Singhal]

Would GLM 4.7 in Cursor!!

[MAiKo26 (Mohamed Aziz Karoui)]

Minimax more so than GLM to be honest

[liquefy]

Yes both GLM and Minimax seem pretty decent for the price. I think it depends how well they will behave within Cursor’s ecosystem and that determines whether they get added. Staff mentioned this a lot in Kimi threads. I hope they are added, but honestly I would appreciate more bug fixes first with stability improvements.

[haris-musa (Haris)]

I don’t think we will get it, we have been crying for this since GLM 4.5 but apparently, they are not interested in these GLM and Minimax models.

They gave us broken Kimi K2, that’s it.

[Eddy]

many issues were fixed with the 4.6 version(for me atleast) am quite satisfied with the 4.7 version in claude code

[Eddy]

Minimax was great for agentic work, especially in terminal!

[Hathor]

I’m using Glm 4.7 + Cursor myself, and I’m happy with it

[Sunny_Anuvadiya]

Lately i think cursor team is ignoring the open soruce models , no news on GLM , minimax. And supprot 20+ modal of openai which are too confusing and cannot be used for their speed.

Would be greate if we get some opensource cost effective modal back in the cursor.

[Daniel_Zolnjan]

Is pasting images into chat working with Glm 4.7?

[Hathor]

It should work, but I have an annual Lite plan, and images work better only with a Pro subscription.

[gustojs (Darek Gusto)]

On GLM Coding Plan inside Cursor, I get “invalid API parameter” error when I send a prompt with pasted image.

So I don’t think it works.

Yes, I’m on a Cursor Pro plan.

[Eddy]

GLM 4.7 itself doesnt support images, its a text only model

if you switch to GLM 4.6V it should (though the 128k context window gets flooded too quickly within IDEs..)

GLM got MCPs which work with their coding plan, maybe try adding them to the mcp.json of cursor so it can call it?

[Hathor]

Pro Glm

[Eddy]

Also just tried and out confirmed

GLM 4.6V does accept images and is able to analyze them (kinda slow though)

4.5 Air doesnt

{“error”:{“type”:“provider”,“reason”:“provider_error”,“message”:“Provider returned 400”,“retryable”:false,“provider”:{“status”:400,“body”:“{“error”:{“code”:“1210”,“message”:“Invalid API parameter, please check the documentation.”}}”}}}

So you can use the Vision models without MCPs right in there

so its not using some internal cursor tool but correclty its own vision capabilties (or lack there off depending on the model)