Skip to content
aiqbench

Zai · model passport

Z.ai: GLM 4.6V

glm-4-6
  • Open weights
  • MIT licence
  • 107.7B params
  • 131,072 context
  • Released 2025-12-07(HF repo created)
13/27axes covered
Overall · LMArena
1,424#113 of 396
Best percentile
96.3BFCL v4 (function calling)
Cohort standing
59.2median 54.2

What the data says

Benchmarked on 13 of 27 tracked axes. Every figure below is the model's own published result on that board.

Strongest axes

  • 96.3 pct — BFCL v4 (function calling) (72.4%)

Uneven: 80.4 percentile points between its best and worst axis — 96.3 pct on BFCL v4 (function calling) versus 15.9 pct on Terminal-Bench.

  • BFCL v4 (function calling)Dormant board96.3 pct72.4%Verify →
  • MMLU-Pro (Artificial Analysis)83.1 pct82.9%Verify →
  • LiveCodeBench (Artificial Analysis)77.1 pct69.5%Verify →
  • Overall71.7 pct1,424Verify →
  • Instruction following71 pct1,413Verify →
  • Hard prompts71 pct1,442Verify →
  • Coding66.5 pct1,458Verify →
  • Factuality Arena62.2 pct1,447Verify →
  • LiveBench60.5 pct59.6%Verify →
  • HHEM Hallucination (Vectara)50.5 pct90.5%Verify →
  • IFBench (Artificial Analysis)48.8 pct43.4%Verify →
  • WebDev Arena22.3 pct1,341Verify →
  • Terminal-Bench15.9 pct24.5%Verify →
  • Vision ArenaNot benchmarked
  • Search ArenaNot benchmarked
  • Document ArenaNot benchmarked
  • Agent ArenaNot benchmarked
  • GPQA DiamondNot benchmarked
  • AIME (math)Not benchmarked
  • FrontierMathNot benchmarked
  • SimpleQA VerifiedNot benchmarked
  • ARC-AGI-2Not benchmarked
  • Humanity's Last ExamNot benchmarked
  • SimpleBenchNot benchmarked
  • SWE-bench Verified (live board)Not benchmarked
  • Terminal-Bench 4.0Not benchmarked
  • SWE-bench VerifiedNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

How to run it

Run it yourself

Repositoryzai-org/GLM-4.6V
Parameters107.7B
ArchitectureGlm4vMoeForConditionalGeneration
LicenceMIT

Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "zai-org/GLM-4.6V"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Ollama

We don't sync Ollama tags, so we won't print a command we can't verify — a wrong one would pull the wrong model. Search the Ollama library →

Unsloth fine-tuning recipes are model-specific and we don't derive them here. Unsloth docs →

Use it hosted

API model IDz-ai/glm-4.6v
Context length131,072 tokens
Max output32,768 tokens
Acceptsimage, text, video
Input price$0.30 per 1M tokens
Cached input price$0.06 per 1M tokens
Output price$0.90 per 1M tokens

API parameters this endpoint supports

Frequency penaltyInclude reasoningMax tokensPresence penaltyReasoningRepetition penaltyResponse formatSeedStop sequencesTemperatureTool choiceTool callingTop-kTop-p

Example request

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"z-ai/glm-4.6v","messages":[{"role":"user","content":"Hello"}]}'

View on OpenRouter →

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.

  • glm-4.6 stands at 59.2 (raw mean 61.3) — measured on 13 of 27 axes.

    standingtemplateVerify →

  • glm-4.6 stands at 59.2 versus a cohort median of 54.2 across 123 models released within 180 days — above the peer median.

    cohorttemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →