Sign Up

GLM 5.3 Flash

Z.ai
Announcement

Cost/1M tokens

input
$0.0399
cached
$0.03
output
$0.0784
Context
1Mtokens
Released
2mo ago
Size
321Bparameters

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Arena leaderboards and other sourced evaluations appear below; missing scores are not zero.

How it ranks

Against its predecessor, similarly priced models and the models ranked around it.

DeepSWE 1.1

Best pass@1 per model · Higher is better

#13 of 28

  1. 74.1%
    GPT-6 Astra
  2. 73.8%
    Gemini 3.8 Flash
  3. 73.6%
    Claude Opus 5
  4. 72.7%
    GPT-5.6 Sol
  5. 69.9%
    Claude Fable 5
  6. 69.6%
    GPT-5.6 Terra
  7. 69.0%
    GLM 5.3
  8. 68.5%
    Kimi K3
  9. 67.5%
    Grok 4.6
  10. 67.2%
    GPT-5.6 Luna
  11. 63.4%
    GLM 5.3 Flash

Arena Text

Best Arena rating per model · Higher is better

Arena #40

  1. 1475
    GLM 5.2Nearby rank
  2. 1475
    DeepSeek V4.1 FlashNearby rank
  3. 1474
    GLM 5.3 Flash
  4. 1474
    GPT-6 AstraNearby rank
  5. 1474
    Claude Opus 4.5Nearby rank
  6. 1318
    gpt-oss-20bSimilar price
  7. 1303
    Gemma 3 4BSimilar price
  8. 1274
    Mistral Small 3Similar price
  9. 1241
    Nova Micro 1.0Similar price

Ahead of 4 of 4 similarly priced models

Arena WebDev

Best Arena rating per model · Higher is better

Arena #35

  1. 1618
    Grok 4.6Nearby rank
  2. 1615
    GPT-5.6 SolNearby rank
  3. 1609
    GLM 5.3 Flash
  4. 1609
    Muse EmberNearby rank
  5. 1608
    GLM 5.2Nearby rank

Arena Vision

Best Arena rating per model · Higher is better

Arena #36

  1. 1277
    GPT-5.5 InstantNearby rank
  2. 1275
    Claude Sonnet 4.6Nearby rank
  3. 1272
    GLM 5.3 Flash
  4. 1266
    GPT-5.6 TerraNearby rank
  5. 1266
    Gemini 3.5 Flash LiteNearby rank

Arena Search

Best Arena rating per model · Higher is better

Not ranked on Arena Search yet

Agent Arena

Best IPS score per model · Higher is better

Arena #39

  1. -0.022
    Gemini 3.7 FlashNearby rank
  2. -0.027
    MiniMax M3.1 FlashNearby rank
  3. -0.029
    GLM 5.3 Flash
  4. -0.037
    Step 5 PreviewNearby rank
  5. -0.045
    Gemini 3.1 Pro PreviewNearby rank

ARC-AGI-2

Best score per model · Higher is better

#28 of 85

  1. 72.1%
    Gemini 3.5 FlashNearby rank
  2. 67.1%
    Grok 4.6Nearby rank
  3. 65.8%
    GLM 5.3 Flash
  4. 65.1%
    Grok 4.20Nearby rank
  5. 64.6%
    Claude Opus 4.6Nearby rank

FrontierMath-Tier-4-v2-Private

Best score per model · Higher is better

#51 of 72

  1. 17.1%
    Grok 4.7Nearby rank
  2. 17.1%
    Grok 4.20 (Reasoning)Nearby rank
  3. 17.1%
    GLM 5.3 Flash
  4. 14.6%
    Grok 4.3Nearby rank
  5. 12.2%
    Kimi K2.7 CodeNearby rank

FrontierMath-Tiers-1-3-v2-Private

Best score per model · Higher is better

#48 of 87

  1. 57.2%
    Grok 4.5Nearby rank
  2. 55.8%
    GPT-5 ProNearby rank
  3. 55.8%
    GLM 5.3 Flash
  4. 55.4%
    GPT-5Nearby rank
  5. 54.0%
    Kimi K2.7 CodeNearby rank
  6. 19.3%
    Qwen3.7 FlashSimilar price

Ahead of 1 of 1 similarly priced model

GPQA diamond

Best score per model · Higher is better

#43 of 161

  1. 90.5%
    GPT-6 LunaNearby rank
  2. 90.2%
    Claude Opus 4.7Nearby rank
  3. 90.2%
    GLM 5.3 Flash
  4. 89.9%
    GLM 5.1Nearby rank
  5. 89.8%
    DeepSeek V4.1 FlashNearby rank
  6. 82.3%
    Qwen3.7 FlashSimilar price
  7. 60.8%
    gpt-oss-20bSimilar price
  8. 23.9%
    Llama 3.2 1B InstructSimilar price
  9. 23.2%
    Gemma 3 4BSimilar price

Ahead of 4 of 4 similarly priced models

OTIS Mock AIME 2024-2025

Best score per model · Higher is better

#46 of 158

  1. 94.4%
    DeepSeek V4 Flash 0731Nearby rank
  2. 94.2%
    Gemini 3.6 FlashNearby rank
  3. 93.9%
    GLM 5.3 Flash
  4. 93.3%
    Qwen3.6 PlusNearby rank
  5. 93.3%
    Qwen3.7 PlusNearby rank
  6. 86.7%
    Qwen3.7 FlashSimilar price
  7. 65.3%
    gpt-oss-20bSimilar price
  8. 7.5%
    Gemma 3 4BSimilar price
  9. 0.6%
    Llama 3.2 1B InstructSimilar price

Ahead of 4 of 4 similarly priced models

DeepSWE 1.1

Primary coding-agent benchmark · 113 tasks · Data generated 2026-09-22

Official leaderboard
63.4%
pass@1 · highest published configuration
mini-swe-agent · max reasoning
pass@4
85.0%
Average cost / attempt
$0.48
Benchmark runs
4

pass@1 is the success rate across scored attempts; pass@4 is the share of tasks solved in at least one of four attempts. Costs are reported by DeepSWE for each configuration. These results are not directly comparable to evaluations using different harnesses. Official source JSON.

Reasoning effort

ARC-AGI-2

Score by reasoning effort · Higher is better

All published scores

Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.

Arena ratings: Arena (lmarena-ai/leaderboard-dataset), licensed CC BY 4.0. Ratings come from pairwise human votes and are only comparable within one leaderboard; the interval is Arena's 95% confidence interval. Model names are matched to the Vertical Tabs catalog.

To give you the best experience, we use cookies for personalization, insights, and performance. Learn more.