Model

GLM-4.7

Recorded as suitable for chat, rag, coding, zh, multilingual.

Model card

Source: Model registry API
Verified: 19 Sep 2026
Confidence: High

01What this is

Author: zai-org
Family: GLM-4
Task: text-generation
Hugging Face: zai-org/GLM-4.7
OpenRouter: z-ai/glm-4.7
Also known as: GLM-4.7, glm 4

02Registry facts

19 Sep 2026

LicenceFACT: mit; FreshRetrieved 19 Sep 2026
ParametersFACT: 358.3B; FreshRetrieved 19 Sep 2026
Context lengthFACT: not returned by the registry; Evidence not verified
Downloads (30 days)FACT: 103,871; FreshRetrieved 19 Sep 2026
LikesFACT: 2,056; FreshRetrieved 19 Sep 2026
GatedFACT: no; FreshRetrieved 19 Sep 2026

Model

downloadsFACT: 103871; Verification may be outdatedRetrieved 19 Sep 2026
gatedFACT: false; Verification may be outdatedRetrieved 19 Sep 2026
hosted context length · AtlasCloudFACT: 202752; Verification may be outdatedRetrieved 19 Sep 2026
hosted context length · VeniceFACT: 198000; Verification may be outdatedRetrieved 19 Sep 2026
hosted context length · DeepInfraFACT: 202752; Verification may be outdatedRetrieved 19 Sep 2026
likesFACT: 2056; Verification may be outdatedRetrieved 19 Sep 2026
model licenseFACT: mit; Verification may be outdatedRetrieved 19 Sep 2026
parameter countFACT: 358337791296; Verification may be outdatedRetrieved 19 Sep 2026

Pricing

hosted price completion per m · AtlasCloudFACT: 1.8500; AgeingRetrieved 19 Sep 2026
hosted price completion per m · VeniceFACT: 1.9292; AgeingRetrieved 19 Sep 2026
hosted price completion per m · DeepInfraFACT: 1.7500; AgeingRetrieved 19 Sep 2026
hosted price prompt per m · AtlasCloudFACT: 0.5200; AgeingRetrieved 19 Sep 2026
hosted price prompt per m · VeniceFACT: 0.4004; AgeingRetrieved 19 Sep 2026
hosted price prompt per m · DeepInfraFACT: 0.4000; AgeingRetrieved 19 Sep 2026

Other

available on platform · vendorFACT

vendor

AgeingRetrieved 19 Sep 2026

available on platform · awsFACT

aws

Evidence not verified — the quoted phrase was not found in the fetched document.

Evidence not verified

last modifiedFACT

2026-01-29T08:05:58.000Z

Verification may be outdatedRetrieved 19 Sep 2026

libraryFACT

transformers

AgeingRetrieved 19 Sep 2026

03VRAM by quantisation

ASSESSMENT

Peak VRAM for serving this model, computed by lib/deployment/hardware.ts from weights + KV cache + 15% runtime overhead. Inputs: 358.3B parameters, 8,192 tokens of context (assumed — none fetched yet), 4 concurrent requests (assumed).

Decimal GB (10⁹ bytes), matching how GPU memory is advertised.
Quantisation	Weights	KV cache	Overhead	Total
bf16bf16 is two bytes per weight, the same arithmetic as fp16.	716.68 GB	15.03 GB	109.76 GB	841.46 GB
fp8	358.34 GB	15.03 GB	56.01 GB	429.38 GB
int4	179.17 GB	15.03 GB	29.13 GB	223.33 GB
q4_k_mSized as 4-bit weights. Q4_K_M averages nearer 4.5 bits in practice, so read this as a floor.	179.17 GB	15.03 GB	29.13 GB	223.33 GB

What this estimate assumes

Weights: 358.3B parameters × 2 bytes per parameter (fp16) = 716.68 GB.
No model configuration was supplied, so the KV cache uses the size heuristic for this band: 0.459 MB per token, from extrapolated above 90B — no configuration was fetched for this band.
KV cache: 0.459 MB per token × 8,192 tokens of context × 4 concurrent requests = 15.03 GB, sized for every user holding a full context at once.
Overhead: 15% of weights plus KV cache for the runtime, activations and memory fragmentation = 109.76 GB.
GB means 10⁹ bytes, matching how GPU memory is advertised.
KV cache held at fp16 (2 bytes per element); quantising weights does not by itself quantise the cache.
Peak figure, not average: prefix caching and paged attention usually keep real usage lower, while speculative decoding, CUDA graphs and a second resident model push it higher.
"Concurrent" counts requests generated at the same instant, not people: 4 simultaneous requests is the worst case this estimate is built on — measure yours before buying.

04Hardware profiles

No hardware profile was sized for this answer.

05Hosted providers

Prompt (per million tokens, USD)FACT

0.39999999999999997

Source: openrouter

FreshRetrieved 19 Sep 2026

Completion (per million tokens, USD)FACT

1.75

FreshRetrieved 19 Sep 2026

Provider	Quantisation	Context
DeepInfra	fp4	202,752
Venice	fp4	198,000
AtlasCloud	fp8	202,752
Novita	fp8	204,800
Google	unknown	200,000
Z.AI	fp4	202,752

06Deployment stacks

0 using it

No published deployment stack names this model yet.

07Evidence

17 verified

01
Tier 2official vendor legal documentationVendor legal and technical documentation
GLM-4.7 - Overview - Z.AI DEVELOPER DOCUMENT
https://docs.z.ai/guides/llm/glm-4.7
AgeingRetrieved 19 Sep 2026sha256:ba793c1b4dc2
1 record
02–10
Tier 4official repository / model cardOpenRouter
OpenRouter endpoints — z-ai/glm-4.7
https://openrouter.ai/api/v1/models/z-ai/glm-4.7/endpoints
Verification may be outdatedRetrieved 19 Sep 2026sha256:1717ccae7f55
9 records · · · · · · · ·
11–17
Tier 4official repository / model cardHugging Face Hub
Hugging Face model card data — zai-org/GLM-4.7
https://huggingface.co/api/models/zai-org/GLM-4.7
Verification may be outdatedRetrieved 19 Sep 2026sha256:a408bfc72b3c
7 records · · · · · ·

Improve this page

From the field

0 deployments · 0 questions

Nobody has reported deploying this here yet, and no question has been opened against this page. Both appear once a reviewer accepts them.