llmboard.aiAI model intelligence
Home

Model Rankings

OverallOpen ModelsAgentCodingReasoningMathKnowledgeInstruction FollowingTextVision
Image GenerationImage Editing
Video GenerationImage to VideoVideo Editing
Text to SpeechSpeech to Text
Embeddings

Efficiency

Chat Token PricingImage PricingVideo PricingAudio Pricing
Chat Speed & LatencyProvider Reliability

Benchmarks

GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-Pro
All Benchmarks

Tools

Model Directory

Scoring & Data

Scoring & Data
1224 models729 benchmarks

Leaderboard Center

Overall RankingCodingCore BenchmarksPrice & ValueRuntime Performance

Modalities

All ModelsImage GenerationImage EditingVideo GenerationImage-to-VideoVideo EditingText-to-SpeechSpeech-to-TextEmbeddings

Data & Methods

Scoring MethodAll Benchmarks
llmboard.aiCopyright 2026 llmboard.ai

reasoning benchmark

Terminal-Bench 2.0 Leaderboard

Terminal-Bench 2.0 is an updated benchmark for evaluating AI agents' ability to use tools and operate a computer through a terminal.

Updated Sep 5, 2026

Models51
Model coverage51
MetricScore
EvidenceC

On this page

  • Ranking
  • Highlights
  • Distribution
  • Top models
  • About
  • FAQ

Terminal-Bench 2.0 Ranking

Higher score ranks better on this benchmark.

30 of 51 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelOPGPT-5.5OpenAIScore82.70%Percentile100.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank02ModelANClaude Mythos PreviewAnthropicScore82.00%Percentile98.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank03ModelANClaude Sonnet 5AnthropicScore80.40%Percentile96.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank04ModelOPGPT-5.3 CodexOpenAIScore77.30%Percentile94.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank05ModelGOGemini 3.5 FlashGoogleScore76.20%Percentile92.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank06ModelOPGPT-5.4OpenAIScore75.10%Percentile90.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank07ModelANClaude Opus 4.8AnthropicScore74.60%Percentile88.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank08ModelACQwen3.7-PlusAlibaba Cloud / Qwen TeamScore70.30%Percentile86.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank09ModelACQwen3.7 MaxAlibaba Cloud / Qwen TeamScore69.70%Percentile84.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank10ModelANClaude Opus 4.7AnthropicScore69.40%Percentile82.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank11ModelZAGLM-5.1Zhipu AIScore69.00%Percentile80.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank12ModelGOGemini 3.1 ProGoogleScore68.50%Percentile78.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank13ModelXIMiMo-V2.5-ProXiaomiScore68.40%Percentile76.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank14ModelDEDeepSeek-V4-Pro-MaxDeepSeekScore67.90%Percentile74.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank15ModelMAKimi K2.6Moonshot AIScore66.70%Percentile72.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank16ModelXIMiMo-V2.5XiaomiScore65.80%Percentile70.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank17ModelANClaude Opus 4.6AnthropicScore65.40%Percentile68.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank18ModelOPGPT-5.2 CodexOpenAIScore64.00%Percentile66.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank19ModelACQwen3.6 PlusAlibaba Cloud / Qwen TeamScore61.60%Percentile64.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank20ModelOPGPT-5.4 miniOpenAIScore60.00%Percentile62.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank21ModelANClaude Opus 4.5AnthropicScore59.30%Percentile60.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank22ModelACQwen3.6-27BAlibaba Cloud / Qwen TeamScore59.30%Percentile58.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank23ModelANClaude Sonnet 4.6AnthropicScore59.10%Percentile56.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank24ModelMEMuse SparkMetaScore59.00%Percentile54.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank25ModelXIMiMo-V2-ProXiaomiScore57.10%Percentile52.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank26ModelMIMiniMax M2.7MiniMaxScore57.00%Percentile50.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank27ModelDEDeepSeek-V4-Flash-MaxDeepSeekScore56.90%Percentile48.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank28ModelDEDeepSeek-V4-Flash-0423DeepSeekScore56.60%Percentile46.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank29ModelZAGLM-5Zhipu AIScore56.20%Percentile44.00%Participants51EvidenceCEvaluatedSep 8, 2026
Rank30ModelMIMAI-Code-1-FlashMicrosoftScore54.80%Percentile42.00%Participants51EvidenceCEvaluatedSep 8, 2026

Terminal-Bench 2.0 Highlights

The leading models and scores on this benchmark.

Terminal-Bench 2.0 Score Distribution

A closer view of the leading scores on this benchmark.

Terminal-Bench 2.0

The Top AI Models for Terminal-Bench 2.0

The first five results on this benchmark, with official price and output speed added where the model identity can be matched.

What is Terminal-Bench 2.0?

What Terminal-Bench 2.0 measures and how its scores work.

Terminal-Bench 2.0 is a benchmark for testing AI agents on real-world, end-to-end tasks performed autonomously through a terminal.

It measures how well models handle tasks including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities, reported as Score in ratio units.

Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of C.

Family
Terminal-Bench 2.0
Modality
text
Primary category
reasoning
Score direction
higher
LLMBoard eligible
No
Evaluation key
overall

LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.

FAQ

Common questions about Terminal-Bench 2.0.

Which model scores highest on Terminal-Bench 2.0?

GPT-5.5 is currently ranked first with 82.70%.

What are the top three models on Terminal-Bench 2.0?

The current leaders are GPT-5.5 (82.70%), Claude Mythos Preview (82.00%), and Claude Sonnet 5 (80.40%).

Which Terminal-Bench 2.0 model has the lowest official input price?

Step-3.5-Flash has the lowest matched official input price at $0.10 input / $0.30 output per 1M tokens.

Which models are fastest among Terminal-Bench 2.0 results?

The fastest matched records are Gemini 3.5 Flash (210.94 tok/s via Google), GPT-5.5 (134.94 tok/s via OpenAI), and Gemini 3 Flash (124.32 tok/s via Google).

Does the highest score result prove overall model quality?

No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.

What does Terminal-Bench 2.0 measure?

It measures how well models handle tasks including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities, reported as Score in ratio units.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

51 model results are currently shown.

Does this benchmark affect the overall score?

No. This benchmark is shown for reference but does not contribute to the overall score.

Rank #1GPT-5.582.70%
Rank #2Claude Mythos Preview82.00%
Rank #3Claude Sonnet 580.40%
Rank #4GPT-5.3 Codex77.30%

Ranking basisThis terminal-bench 2.0 AI model leaderboard uses descending score in the benchmark original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.

  1. 01
    OP
    GPT-5.5OpenAI
    Score
    82.70%
    Price
    $5.0 input / $30 output per 1M tokens
    Speed
    Up to 134.94 tok/s via OpenAI

    Strengths

    • Ranks #1 of 51 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures Terminal-Bench 2.0, not total model capability
  2. 02
    AN
    Claude Mythos PreviewAnthropic
    Score
    82.00%

    Strengths

    • Ranks #2 of 51 compared models
    • 98th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures Terminal-Bench 2.0, not total model capability
  3. 03
    AN
    Anthropic
    Score
    80.40%
    Price
    $2.0 input / $10 output per 1M tokens
    Speed
    Up to 42.00 tok/s via Anthropic

    Strengths

    • Ranks #3 of 51 compared models
    • 96th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures Terminal-Bench 2.0, not total model capability
  4. 04
    OP
    OpenAI
    Score
    77.30%
    Price
    $1.8 input / $14 output per 1M tokens
    Speed
    Up to 50.00 tok/s via OpenAI

    Strengths

    • Ranks #4 of 51 compared models
    • 94th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures Terminal-Bench 2.0, not total model capability
  5. 05
    GO
    Google
    Score
    76.20%
    Price
    $1.5 input / $9.0 output per 1M tokens
    Speed
    Up to 210.94 tok/s via Google

    Strengths

    • Ranks #5 of 51 compared models
    • 92th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures Terminal-Bench 2.0, not total model capability

Selection summary

Best AI Models for Terminal-Bench 2.0

GPT-5.5 currently leads Terminal-Bench 2.0 with 82.70%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price, and runtime.

Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.

Claude Sonnet 5
GPT-5.3 Codex
Gemini 3.5 Flash
Benchmark rank #1GPT-5.582.70% · $5.0 input / $30 output per 1M tokens
Benchmark rank #2Claude Mythos Preview82.00%
Benchmark rank #3Claude Sonnet 580.40% · $2.0 input / $10 output per 1M tokens