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 DirectoryCompare Models

Scoring & Data

Scoring & Data
393 models668 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 BenchmarksReasoningMath

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai

reasoning benchmark

Terminal-Bench 2.0 Leaderboard

Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.

Updated Aug 17, 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-5.5OpenAIScore82.7%Percentile100.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank02ModelANClaude Mythos PreviewAnthropicScore82.0%Percentile98.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank03ModelANClaude Sonnet 5AnthropicScore80.4%Percentile96.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank04ModelOPGPT-5.3 CodexOpenAIScore77.3%Percentile94.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank05ModelGOGemini 3.5 FlashGoogleScore76.2%Percentile92.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank06ModelOPGPT-5.4OpenAIScore75.1%Percentile90.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank07ModelANClaude Opus 4.8AnthropicScore74.6%Percentile88.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank08ModelACQwen3.7-PlusAlibaba Cloud / Qwen TeamScore70.3%Percentile86.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank09ModelACQwen3.7 MaxAlibaba Cloud / Qwen TeamScore69.7%Percentile84.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank10ModelANClaude Opus 4.7AnthropicScore69.4%Percentile82.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank11ModelZAGLM-5.1Zhipu AIScore69.0%Percentile80.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank12ModelGOGemini 3.1 ProGoogleScore68.5%Percentile78.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank13ModelXIMiMo-V2.5-ProXiaomiScore68.4%Percentile76.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank14ModelDEDeepSeek-V4-Pro-MaxDeepSeekScore67.9%Percentile74.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank15ModelMAKimi K2.6Moonshot AIScore66.7%Percentile72.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank16ModelXIMiMo-V2.5XiaomiScore65.8%Percentile70.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank17ModelANClaude Opus 4.6AnthropicScore65.4%Percentile68.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank18ModelOPGPT-5.2 CodexOpenAIScore64.0%Percentile66.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank19ModelACQwen3.6 PlusAlibaba Cloud / Qwen TeamScore61.6%Percentile64.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank20ModelOPGPT-5.4 miniOpenAIScore60.0%Percentile62.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank21ModelANClaude Opus 4.5AnthropicScore59.3%Percentile60.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank22ModelACQwen3.6-27BAlibaba Cloud / Qwen TeamScore59.3%Percentile58.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank23ModelANClaude Sonnet 4.6AnthropicScore59.1%Percentile56.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank24ModelMEMuse SparkMetaScore59.0%Percentile54.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank25ModelXIMiMo-V2-ProXiaomiScore57.1%Percentile52.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank26ModelMIMiniMax M2.7MiniMaxScore57.0%Percentile50.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank27ModelDEDeepSeek-V4-Flash-MaxDeepSeekScore56.9%Percentile48.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank28ModelDEDeepSeek-V4-Flash-0423DeepSeekScore56.6%Percentile46.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank29ModelZAGLM-5Zhipu AIScore56.2%Percentile44.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank30ModelMIMAI-Code-1-FlashMicrosoftScore54.8%Percentile42.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank31ModelGOGemini 3 ProGoogleScore54.2%Percentile40.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank32ModelOPGPT-5.1 CodexOpenAIScore52.8%Percentile38.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank33ModelACQwen3.5-397B-A17BAlibaba Cloud / Qwen TeamScore52.5%Percentile36.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank34ModelACQwen3.6-35B-A3BAlibaba Cloud / Qwen TeamScore51.5%Percentile34.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank35ModelSTStep-3.5-FlashStepFunScore51.0%Percentile32.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank36ModelMAKimi K2.5Moonshot AIScore50.8%Percentile30.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank37ModelACQwen3.5-122B-A10BAlibaba Cloud / Qwen TeamScore49.4%Percentile28.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank38ModelGOGemini 3 FlashGoogleScore47.6%Percentile26.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank39ModelDEDeepSeek-V3.2 (Thinking)DeepSeekScore46.4%Percentile24.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank40ModelDEDeepSeek-V3.2DeepSeekScore46.4%Percentile22.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank41ModelDEDeepSeek-V3.2-SpecialeDeepSeekScore46.4%Percentile20.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank42ModelOPGPT-5.4 nanoOpenAIScore46.3%Percentile18.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank43ModelMIMAI-Thinking-1MicrosoftScore46.0%Percentile16.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank44ModelACQwen3.5-27BAlibaba Cloud / Qwen TeamScore41.6%Percentile14.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank45ModelZAGLM-4.7Zhipu AIScore41.0%Percentile12.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank46ModelACQwen3.5-35B-A3BAlibaba Cloud / Qwen TeamScore40.5%Percentile10.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank47ModelXIMiMo-V2-FlashXiaomiScore38.5%Percentile8.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank48ModelPOLaguna XS 2.1PoolsideScore37.5%Percentile6.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank49ModelACQwen3-Coder 480B A35B InstructAlibaba Cloud / Qwen TeamScore37.5%Percentile4.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank50ModelCONorth Mini Code 1.0CohereScore36.0%Percentile2.0%Participants51EvidenceCEvaluatedAug 17, 2026
Rank51ModelNVNemotron 3 Super (120B A12B)NVIDIAScore31.0%Percentile0.0%Participants51EvidenceCEvaluatedAug 17, 2026

Terminal-Bench 2.0 Highlights

The leading models and scores on this benchmark.

Rank #1GPT-5.582.7%Rank #2Claude Mythos Preview82.0%Rank #3Claude Sonnet 580.4%Rank #4GPT-5.3 Codex77.3%

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.

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

  1. 01
    OP
    GPT-5.5OpenAI
    Score
    82.7%
    Price
    $5.0 input / $30 output per 1M tokens

    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
    82.0%

    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
    Claude Sonnet 5Anthropic
    Score
    80.4%
    Price
    $2.0 input / $10 output per 1M tokens
    Speed
    Up to 11 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
    GPT-5.3 CodexOpenAI
    Score
    77.3%
    Price
    $1.8 input / $14 output per 1M tokens
    Speed
    Up to 2.4 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
    Gemini 3.5 FlashGoogle
    Score
    76.2%
    Price
    $1.5 input / $9.0 output per 1M tokens
    Speed
    Up to 1.2 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 82.7%. 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.

Benchmark rank #1GPT-5.582.7% · $5.0 input / $30 output per 1M tokensBenchmark rank #2Claude Mythos Preview82.0%Benchmark rank #3Claude Sonnet 580.4% · $2.0 input / $10 output per 1M tokens

What is Terminal-Bench 2.0?

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

Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.

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
Yes
Evaluation key
terminal-bench-2|llm-stats-current

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 82.7%.

What does Terminal-Bench 2.0 measure?

Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.

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?

Yes. This benchmark can contribute to the current LLMBoard capability score.