Skip to main content

Model profile

GLM-5.1

Z.AISupersededReleased Apr 7, 2026
Data verified
Superseded:Z.AI has released newer models in this line —GLM-5.2
Overall Score
67.74Public #18 of 200Verified #16 of 99
Arena Elo
1470
Eligible category ranks
3of 8
Price (1M tokens)
$1.4 in / $4.4 out
API pricing
Speed
Not listed
Context
203K

Evidence coverage

36 of 323 tracked benchmarks are published. 20 are verified and 16 provisional. 7 of 8 categories are measured.

Updated July 23, 2026Methodology
Published / tracked
36 / 323
Verified
20
Provisional
16
Categories with evidence
7 / 8

Evidence by category

  • Agentic12 benchmarks
    Mixed evidence
  • Coding6 benchmarks
    Mixed evidence
  • Reasoning2 benchmarks
    Reported
  • Knowledge8 benchmarks
    Mixed evidence
  • Math6 benchmarks
    Verified
  • Multilingual0 benchmarks
    Not measured
  • Multimodal1 benchmark
    Reported
  • Inst. Following1 benchmark
    Reported
Open WeightSelf-hostReasoning
Confidence:
Medium
snapshot

GLM-5.1 ranks #18 out of 200 models on the public leaderboard with an overall score of 67.74/100. It also ranks #16 out of 99 on the verified leaderboard. This places it in the mid-tier of AI models, with strengths in specific benchmark categories.

GLM-5.1 is a open weight model with a 203K token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.

GLM-5.1 sits inside the GLM-5 family alongside GLM-5, GLM-5.2, GLM-5 (Reasoning), GLM-5-Turbo, GLM-5V-Turbo. BenchLM links it directly to GLM-5 as the earlier related model in that lineage. This profile currently has 36 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Mathematics (#3), while its weakest is Agentic (#54). This performance profile makes it particularly strong for mathematical reasoning, scientific computing, and quantitative analysis.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 67.0167.78

  1. MiMo-V2-Pro
    Xiaomi
    #1767.78
    MiMo-V2-Pro is #17 with a score of 67.78.
    Compare
  2. GLM-5.1Current model
    Z.AI
    #1867.74
    GLM-5.1 is #18 with a score of 67.74.
  3. Gemini 3 Pro
    Google
    #1967.73
    Gemini 3 Pro is #19 with a score of 67.73.
    Compare
  4. Inkling
    Thinking Machines Lab
    #2067.54
    Inkling is #20 with a score of 67.54.
    Compare
  5. Qwen3.7 Plus
    Alibaba
    #2167.22
    Qwen3.7 Plus is #21 with a score of 67.22.
    Compare
  6. GPT-5.6 Luna
    OpenAI
    #2267.17
    GPT-5.6 Luna is #22 with a score of 67.17.
    Compare
  7. GPT-5.2 Pro
    OpenAI
    #2367.01
    GPT-5.2 Pro is #23 with a score of 67.01.
    Compare

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Math67%
    Eligible cohort rank #3 of 7Category score 64.6
  2. Coding76%
    Eligible cohort rank #30 of 122Category score 57.5
  3. Agentic55%
    Eligible cohort rank #54 of 119Category score 48.6

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #54 of 119Percentile 55thWeight 22%12 benchmarksMixed sources48.6
CodingRank #30 of 122Percentile 76thWeight 20%6 benchmarksMixed sources57.5
ReasoningRank Not rankedWeight 17%2 benchmarksReported66.3
KnowledgeRank Not rankedWeight 12%8 benchmarksMixed sources82.3
MathRank #3 of 7Percentile 67thWeight 5%6 benchmarksVerified64.6
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalWeight 12%1 benchmarkReportedScore pending
Inst. FollowingRank Not rankedWeight 5%1 benchmarkReported92.6

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

GLM-5.1
API / mo$4,350
Self-host / mo$18,221
Break-even264M/day
Model the full break-even

Chatbot Arena performance

Scroll horizontally to inspect confidence intervals and vote counts.

Chatbot Arena Elo, confidence interval, and vote count by evaluation view
ViewEloConfidence intervalVotes
Text Overall1470±4.629,919
Coding1521±7.48,422
Math1481±15.31,559
Instruction Following1464±6.810,098
Creative Writing1453±9.25,126
Multi-turn1483±9.24,945
Hard Prompts1492±5.519,738
Hard Prompts (English)1503±7.09,647
Longer Query1484±6.513,101

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic12 benchmarks
Terminal-Bench 2.0Provider exact
63.5%Weighted 38%
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
BrowseCompProvider exact
68%Weighted 28%
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
τ³-bench resultsProvider exact

τ³-Bench Tool-Agent-User Evaluation

70.6%Display only
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
MCP AtlasProvider exact
71.8%Display only
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
CyberGymBenchmark exact
68.7%Display only
Source: CyberGym leaderboardProvenance: CyberGym Level 1 reports Claude Code with GLM-5.1 on the public leaderboard. BenchLM stores the reported target-vulnerability reproduction success rate on the local cyberGym key.
Claw-EvalBenchmark exact
62.3%Display only
Source: Claw-Eval leaderboardProvenance: Claw-Eval reports this model as glm51 in the official 2026-05-09 leaderboard snapshot. BenchLM stores the primary Pass^3 value on the local Claw-Eval display key.
AA Agentic IndexReported

Artificial Analysis Agentic Index

29.9%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

97.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
GDPval-AAReported

GDPval-AA normalized

37.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Gert LabsBenchmark exact

Gert Labs Composite Game Benchmark

60.11%Display only
Source: Gert Labs rankingsProvenance: Gert Labs reports this composite leaderboard score in the public rankings API. BenchLM scales the source gscore from 0-1 to 0-100 and stores it as a display-only agentic benchmark.
GDPval-AAReported
1257Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
ResearchClawBenchBenchmark exact
18.2%Display only
Source: ResearchClawBench leaderboardProvenance: ResearchClawBench reports this model as ResearchHarness (GLM-5.1) in the official Pass@1 leaderboard. BenchLM stores the one-decimal RADS average on the local ResearchClawBench display key and excludes it from weighted rankings.
Coding6 benchmarks
SWE-RebenchBenchmark exact
62.7%Weighted 20%
Source: SWE-Rebench leaderboardProvenance: Live default SWE-Rebench leaderboard lists GLM-5.1 at 62.7% resolved rate.
SWE-bench ProProvider exact
58.4%Weighted 10%
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
NL2RepoProvider exact
42.7%Display only
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
Vibe Code BenchBenchmark exact

Vibe Code Bench v1.1

31.46%Display only
Source: Vals AI: Vibe Code Bench v1.1Provenance: Vals Vibe Code Bench v1.1 reports this exact row under zai/glm-5.1-thinking; BenchLM stores it on the local vibeCodeBench key.
AA Coding IndexReported

Artificial Analysis Coding Index

55.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-SciCodeReported

Artificial Analysis SciCode

43.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Reasoning2 benchmarks
AA-LCRReported

Artificial Analysis Long Context Reasoning

62.3%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
CritPtReported

Critical Physics Tasks

4.6%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Knowledge8 benchmarks
HLEProvider exact

Humanity's Last Exam

52.3%Weighted 45%
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Z.AI reports GLM-5.1 exact scores for reasoning and agentic tasks. HLE score of 52.3 represents tool-assisted performance.
GPQA-DProvider exact

GPQA Diamond

86.2%Display only
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
Artificial Analysis Intelligence IndexReported
40.2%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-GPQA DiamondReported

Artificial Analysis GPQA Diamond

86.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-HLEReported

Artificial Analysis Humanity's Last Exam

28.0%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience IndexReported

Artificial Analysis Omniscience Index

1.9%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience AccuracyReported

Artificial Analysis Omniscience Accuracy

24.2%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience Hallucination RateReported

Artificial Analysis Omniscience Hallucination Rate

29.4%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
Math6 benchmarks
FrontierMath v2 (Tiers 1-3)Benchmark exact

FrontierMath v2 Tiers 1-3

33.448%Weighted 30%
Source: Epoch AI FrontierMath v2 leaderboardProvenance: Epoch AI reports FrontierMath v2 Tiers 1-3 at 33.448% for glm-5.1. BenchLM selects the highest published thinking effort for the model and stores the v2 benchmark slice separately.
AIME26Provider exact

AIME 2026

95.3%Weighted 25%
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
HMMT Feb 2026Provider exact

Harvard-MIT Mathematics Tournament February 2026

82.6%Weighted 25%
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
FrontierMath v2 (Tier 4)Benchmark exact

FrontierMath v2 Tier 4

12.500%Weighted 10%
Source: Epoch AI FrontierMath v2 leaderboardProvenance: Epoch AI reports FrontierMath v2 Tier 4 at 12.5% for glm-5.1. BenchLM selects the highest published thinking effort for the model and stores the v2 benchmark slice separately.
HMMT Nov 2025Provider exact

Harvard-MIT Mathematics Tournament November 2025

94.0%Display only
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
MMAnswerBenchProvider exact
83.8%Display only
Source: Z.AI: GLM-5.1 Towards Long-Horizon TasksProvenance: Provider exact
Multimodal1 benchmark
Design Arena WebsiteReported

Design Arena Website Elo

1305Display only
Source: OpenRouter model benchmarksProvenance: Display-only Design Arena Website Elo synced from OpenRouter model benchmark metadata. It is excluded from BenchLM weighted scoring.
Inst. Following1 benchmark
AA-IFBenchReported

Artificial Analysis IFBench

76.3%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.

Frequently Asked Questions

How does GLM-5.1 perform overall in AI benchmarks?

GLM-5.1 currently ranks #18 out of 200 models on BenchLM's provisional leaderboard with an overall score of 67.74. It also ranks #16 out of 99 on the verified leaderboard. It is created by Z.AI. Its published context window is 203K.

Is GLM-5.1 good for knowledge and understanding?

GLM-5.1 has visible benchmark coverage in knowledge and understanding, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.1 good for coding and programming?

GLM-5.1 ranks #30 out of 122 models in coding and programming benchmarks with an average score of 57.5. There are stronger options in this category.

Is GLM-5.1 good for mathematics?

GLM-5.1 ranks #3 out of 7 models in mathematics benchmarks with an average score of 64.6. It is among the top performers in this category.

Is GLM-5.1 good for reasoning and logic?

GLM-5.1 has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.1 good for agentic tool use and computer tasks?

GLM-5.1 ranks #54 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 48.6. There are stronger options in this category.

Is GLM-5.1 good for multimodal and grounded tasks?

GLM-5.1 has visible benchmark coverage in multimodal and grounded tasks, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.1 good for instruction following?

GLM-5.1 has visible benchmark coverage in instruction following, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.1 open source?

Yes, GLM-5.1 is an open weight model created by Z.AI, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

Which sibling models are related to GLM-5.1?

GLM-5.1 belongs to the GLM-5 family. Related variants on BenchLM include GLM-5, GLM-5.2, GLM-5 (Reasoning), GLM-5-Turbo, GLM-5V-Turbo.

Does GLM-5.1 have full benchmark coverage on BenchLM?

Not yet. GLM-5.1 currently has 36 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of GLM-5.1?

GLM-5.1 has a published context window of 203K, which determines how much text it can process in a single interaction.

Last updated: July 23, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.