VerdictPal
Join

Models / Zhipu

GLM-5.1

GLM · Released 2026-04-07

Z.ai's April 2026 flagship for long-horizon agentic engineering. Official docs list text input/output, 200K context, 128K maximum output, function calling, structured output, context caching, MCP, and open weights under the MIT license.

Long-horizon coding and autonomous agent model for Claude Code/OpenClaw-style workflows when you want open weights from the GLM line.

#25 of 51 on AA Index · current snapshot
40 AA Index · 10 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite52
Context200K
Input / 1M—
Output / 1M—
Knowledge cutoffopen
Output speed85 t/s
TTFT (default API)0.86s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
87
Coding
44
Agentic
62
Long context
62

64average across 4 tested families

Editor's note
GLM-5.1 is the GLM card to watch for long autonomous coding runs. Z.ai's own benchmark claims are strong, but we keep the card at solid until independent score rows are imported into VerdictPal's ledger.
Benchmark placements

Where GLM-5.1 places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
IFBenchReasoning#12/ 5776Artificial Analysis2026-07-21
AA output speedPerformance#32/ 69Artificial Analysis2026-07-21
Artificial Analysis Intelligence IndexReasoning#32/ 6740Artificial Analysis2026-07-21
Terminal-BenchAgentic#35/ 6962Artificial Analysis2026-07-21
GPQA DiamondReasoning#37/ 6987Artificial Analysis2026-07-21
Humanity's Last ExamReasoning#39/ 6528Artificial Analysis2026-07-21
AA-LCRLong context#44/ 6462Artificial Analysis2026-07-21
SciCodeCoding#44/ 6544Artificial Analysis2026-07-21
AA time to first tokenPerformance#48/ 69Artificial Analysis2026-07-21
τ³-Bench BankingAgentic#54/ 6412Artificial Analysis2026-07-21
Family context

The three highest-scoring carded models in each capability family. Where GLM-5.1 shows up, it's highlighted.

ReasoningGLM-5.1 · 87
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingGLM-5.1 · 44
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticGLM-5.1 · 62
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextGLM-5.1 · 62
  1. 1
    Kimi K3
    75
  2. 2
    GPT-5.1
    75
  3. 3
    GPT-5.6 Sol
    74
Newest receipts

Benchmark rows added to the public ledger for GLM-5.1 in the last 120 days. Older rows live in the full table below.

10 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context62.33%Artificial Analysis2
AA output speedPerformance85 t/sArtificial Analysis2
AA time to first tokenPerformance0.86sArtificial Analysis2
Artificial Analysis Intelligence IndexReasoning40.2%Artificial Analysis2
GPQA DiamondReasoning86.8%Artificial Analysis2
Humanity's Last ExamReasoning28%Artificial Analysis2
IFBenchReasoning76.26%Artificial Analysis2
SciCodeCoding43.8%Artificial Analysis2
τ³-Bench BankingAgentic11.55%Artificial Analysis2
Terminal-BenchAgentic61.8%Artificial Analysis2
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning87
Coding44
Agentic62
Long context62
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMiniMaxxAIMetaZhipuXiaomiMeituanCohere

Full benchmark ledger

Every catalog benchmark for GLM-5.1. Scores link to the original source; gaps mean no public row exists yet.

10 sourced rows

10 of 32 catalog benchmarks have a sourced row for GLM-5.1.

Reasoning

Coding

Agentic

Long context

Performance

Changelog
  1. Released GLM-5.1 with 200K context, 128K max output, long-horizon coding focus, and MIT-licensed open weights.
Sources