Inference
EN
Z.AI

GLM-4.7

GLM-4.7 is a reasoning model from Z.AI. 27 benchmarks count toward its index, in 6 categories.

  • Open weights
  • Reasoning
  • Input: text
  • Released 2025-12
RankedEnough results for an index.
IndexAll categories, out of 100.50.0±4.6Median 51.0
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.28/sMedian 67/s
First tokenSeconds to the first output token.1.1 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.60Median $0.60
Output / 1MUS dollars per 1M output tokens.$2.20Median $2.55
ContextMaximum tokens in one request.205KMedian 262K
EloLMArena rating and rank.1435±6 #89Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
z-ai/glm-4.7
Max outputMaximum output tokens per request.
131K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.11
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
6 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Deep InfraHost$0.40$1.75203K
Zhipu AIMaker$0.448$2.09205K
Amazon BedrockDeprecatedCloud$0.60$2.20203K
Novita AIHost$0.60$2.20205K
OpenRouterRouter$0.60$2.20205K
Z.AIMaker$0.60$2.20205K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Run locally

358.34B parameters
16-bit
753.5 GB
8-bit
400.8 GB
4-bit
229.1 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.8 counted48.8138 / 218
CodingCode writing and repair.4 counted53.2115 / 209
ReasoningLogic problems and puzzles.3 counted48.7126 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.5 counted50.9131 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted52.5—
MathMath problems.6 counted49.190 / 168

Results

27 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
τ²-Bench Tool-Agent-User EvaluationAgentic95.9%13 / 149
Harvard-MIT Mathematics Tournament February 2025Math97.1%2 / 15
American Invitational Mathematics Examination 2025Math95.7%8 / 53
AIMEMath93.3% ±0.415 / 95
Artificial Analysis LiveCodeBenchCoding89.4%2 / 12
Terminal-Bench 1.0Agentic50.0% ±5.610 / 47
Massive Multitask Language Understanding ProfessionalKnowledge84.3%25 / 75
Artificial Analysis GPQA DiamondKnowledge85.9%70 / 199
GPQA diamond · Epoch AIKnowledge83.3% ±2.477 / 217
LiveCodeBench v6Coding84.9%21 / 59
Artificial Analysis IFBenchInstruction67.9%52 / 146
Graduate-Level Google-Proof Q&AKnowledge85.7%51 / 141
LiveCodeBenchCoding82.2% ±1.053 / 142
OTIS Mock AIME 2024-2025Math83.3% ±4.879 / 200
Artificial Analysis Humanity's Last ExamKnowledge27.4%94 / 222
Software Engineering Benchmark VerifiedCoding73.8%51 / 114
Artificial Analysis Omniscience AccuracyKnowledge29.3%80 / 164
Artificial Analysis Long Context ReasoningReasoning71.0%108 / 217
GPQA Diamond · Vals AIKnowledge80.0% ±2.068 / 135
Critical Physics TasksReasoning1.7%109 / 214
GDPval-AA normalizedAgentic25.7%70 / 133
Terminal-Bench 2.0 · Vals AIAgentic38.2% ±5.236 / 66
MMLU ProKnowledge82.7% ±0.476 / 135
IOI v1Coding7.6% ±3.438 / 61
Chess PuzzlesReasoning21.0% ±4.159 / 90
Artificial Analysis MMLU-ProKnowledge85.6%10 / 15
Gert Labs Composite Game BenchmarkAgentic40.0%37 / 52
SimpleQA VerifiedKnowledge32.2% ±1.566 / 91
SWE-benchCoding69.4% ±2.166 / 87
Humanity's Last ExamKnowledge24.8%59 / 77
FrontierMath-2025-02-28-PrivateMath2.4% ±0.953 / 69
FrontierMath-Tier-4-2025-07-01-PrivateMath0.0%45 / 55
Terminal-Bench 2.0 · BenchLMAgentic41.0%37 / 41
BrowseCompAgentic52.0%57 / 63
VITA-BenchAgentic15.5%13 / 13
MedQAHealthcare93.7% ±0.221 / 95
LegalBenchLegal83.4% ±0.555 / 149
CaseLaw v2Legal54.9% ±0.936 / 53
TaxEval v2Finance68.8% ±0.999 / 143
Finance Agent v1.1Finance46.0% ±2.836 / 51
MGSMMultilingual88.2% ±1.853 / 74
MedCodeHealthcare32.8% ±2.086 / 106
CorpFin v2Finance46.4% ±0.9113 / 131
MedScribeHealthcare68.6% ±2.198 / 109
Artificial Analysis AIME 2025Math95.0%—
SWE-RebenchCoding58.7%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedEpoch AICC BYmodels.devMIT

History

5 releases
Index51.0 → 50.0
23 Oct 202612 Nov 2026
Output / 1M$1.98 → $2.20
23 Oct 202612 Nov 2026

Questions

How much does GLM-4.7 cost?
$0.60 per 1M input tokens, $2.20 per 1M output tokens.
How good is GLM-4.7?
Its index is 50.0 out of 100. That is place 150 of 283 ranked models.
What is GLM-4.7 best at?
Knowledge: 50.9 out of 100, place 131 of 268.
How fast is GLM-4.7?
28 output tokens per second, 1.1 s to the first token.
What is the context window of GLM-4.7?
205K tokens.
Is GLM-4.7 open source?
Yes. Its weights are published.
When was GLM-4.7 released?
December 2025.

Badge

GLM-4.7 on Inference360: place 150 of 283, index 50.0
HTML
<a href="https://inference360.ai/models/glm-4-7"><img src="https://inference360.ai/badges/glm-4-7.svg" alt="GLM-4.7 on Inference360: place 150 of 283, index 50.0" height="20"></a>
Markdown
[![GLM-4.7 on Inference360: place 150 of 283, index 50.0](https://inference360.ai/badges/glm-4-7.svg)](https://inference360.ai/models/glm-4-7)

Compare with

Grok 450.0MiniMax M2.550.2Mistral Medium 3.5 128B50.3Gemini 3.1 Flash Lite Preview49.6GLM-551.6

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from Z.AI

All models
GLM 4.6VUnrankedGLM-4.7-FlashUnrankedGLM-551.6GLM-5 (Reasoning)UnrankedGLM-5-Turbo51.9