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

SWE-Bench Verified Leaderboard

SWE-Bench Verified is a verified subset of 500 software engineering problems from real GitHub issues.

Updated Sep 5, 2026

Models100
Model coverage113
MetricScore
EvidenceC

On this page

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

SWE-Bench Verified Ranking

Higher score ranks better on this benchmark.

30 of 100 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelANClaude Fable 5AnthropicScore95.00%Percentile100.00%Participants113EvidenceCEvaluatedSep 8, 2026
Rank02ModelANClaude Mythos PreviewAnthropicScore93.90%Percentile99.11%Participants113EvidenceCEvaluatedSep 8, 2026
Rank03ModelANClaude Opus 4.8AnthropicScore88.60%Percentile98.21%Participants113EvidenceCEvaluatedSep 8, 2026
Rank04ModelANClaude Opus 4.7AnthropicScore87.60%Percentile97.32%Participants113EvidenceCEvaluatedSep 8, 2026
Rank05ModelANClaude Sonnet 5AnthropicScore85.20%Percentile96.43%Participants113EvidenceCEvaluatedSep 8, 2026
Rank06ModelANClaude Opus 4.5AnthropicScore80.90%Percentile95.54%Participants113EvidenceCEvaluatedSep 8, 2026
Rank07ModelANClaude Opus 4.6AnthropicScore80.80%Percentile94.64%Participants113EvidenceCEvaluatedSep 8, 2026
Rank08ModelDEDeepSeek-V4-Pro-MaxDeepSeekScore80.60%Percentile93.75%Participants113EvidenceCEvaluatedSep 8, 2026
Rank09ModelGOGemini 3.1 ProGoogleScore80.60%Percentile92.86%Participants113EvidenceCEvaluatedSep 8, 2026
Rank10ModelMIMiniMax M3MiniMaxScore80.50%Percentile91.96%Participants113EvidenceCEvaluatedSep 8, 2026
Rank11ModelACQwen3.7 MaxAlibaba Cloud / Qwen TeamScore80.40%Percentile91.07%Participants113EvidenceCEvaluatedSep 8, 2026
Rank12ModelTMInkling-SmallThinking Machines LabScore80.20%Percentile90.18%Participants113EvidenceCEvaluatedSep 8, 2026
Rank13ModelMAKimi K2.6Moonshot AIScore80.20%Percentile89.29%Participants113EvidenceCEvaluatedSep 8, 2026
Rank14ModelMIMiniMax M2.5MiniMaxScore80.20%Percentile88.39%Participants113EvidenceCEvaluatedSep 8, 2026
Rank15ModelOPGPT-5.2OpenAIScore80.00%Percentile87.50%Participants113EvidenceCEvaluatedSep 8, 2026
Rank16ModelANClaude Sonnet 4.6AnthropicScore79.60%Percentile86.61%Participants113EvidenceCEvaluatedSep 8, 2026
Rank17ModelDEDeepSeek-V4-Flash-MaxDeepSeekScore79.00%Percentile85.71%Participants113EvidenceCEvaluatedSep 8, 2026
Rank18ModelXIMiMo-V2.5-ProXiaomiScore78.90%Percentile84.82%Participants113EvidenceCEvaluatedSep 8, 2026
Rank19ModelACQwen3.6 PlusAlibaba Cloud / Qwen TeamScore78.80%Percentile83.93%Participants113EvidenceCEvaluatedSep 8, 2026
Rank20ModelDEDeepSeek-V4-Flash-0423DeepSeekScore78.60%Percentile83.04%Participants113EvidenceCEvaluatedSep 8, 2026
Rank21ModelGOGemini 3 FlashGoogleScore78.00%Percentile82.14%Participants113EvidenceCEvaluatedSep 8, 2026
Rank22ModelTEHy3TencentScore78.00%Percentile81.25%Participants113EvidenceCEvaluatedSep 8, 2026
Rank23ModelXIMiMo-V2-ProXiaomiScore78.00%Percentile80.36%Participants113EvidenceCEvaluatedSep 8, 2026
Rank24ModelZAGLM-5Zhipu AIScore77.80%Percentile79.46%Participants113EvidenceCEvaluatedSep 8, 2026
Rank25ModelACQwen3.7-PlusAlibaba Cloud / Qwen TeamScore77.70%Percentile78.57%Participants113EvidenceCEvaluatedSep 8, 2026
Rank26ModelMAMistral Medium 3.5Mistral AIScore77.60%Percentile77.68%Participants113EvidenceCEvaluatedSep 8, 2026
Rank27ModelMEMuse SparkMetaScore77.40%Percentile76.79%Participants113EvidenceCEvaluatedSep 8, 2026
Rank28ModelACQwen3.6-27BAlibaba Cloud / Qwen TeamScore77.20%Percentile75.89%Participants113EvidenceCEvaluatedSep 8, 2026
Rank29ModelMAKimi K2.5Moonshot AIScore76.80%Percentile75.00%Participants113EvidenceCEvaluatedSep 8, 2026
Rank30ModelBYSeed 2.0 ProByteDanceScore76.50%Percentile74.11%Participants113EvidenceCEvaluatedSep 8, 2026

SWE-Bench Verified Highlights

The leading models and scores on this benchmark.

SWE-Bench Verified Score Distribution

A closer view of the leading scores on this benchmark.

SWE-Bench Verified

The Top AI Models for SWE-Bench Verified

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

What is SWE-Bench Verified?

What SWE-Bench Verified measures and how its scores work.

SWE-Bench Verified is a benchmark of software engineering problems validated by human annotators for evaluating language models on real-world coding issues.

It measures language models' ability to resolve coding issues by generating patches for Python codebases, using Score as a ratio.

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

Family
SWE-Bench Verified
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 SWE-Bench Verified.

Which model scores highest on SWE-Bench Verified?

Claude Fable 5 is currently ranked first with 95.00%.

What are the top three models on SWE-Bench Verified?

The current leaders are Claude Fable 5 (95.00%), Claude Mythos Preview (93.90%), and Claude Opus 4.8 (88.60%).

Which SWE-Bench Verified 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 SWE-Bench Verified results?

The fastest matched records are Devstral Medium (137.10 tok/s via Mistral AI), Devstral Small 1.1 (137.10 tok/s via Mistral AI), 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 SWE-Bench Verified measure?

It measures language models' ability to resolve coding issues by generating patches for Python codebases, using Score as a ratio.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

100 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 #1Claude Fable 595.00%
Rank #2Claude Mythos Preview93.90%
Rank #3Claude Opus 4.888.60%
Rank #4Claude Opus 4.787.60%

Ranking basisThis swe-bench verified 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
    AN
    Claude Fable 5Anthropic
    Score
    95.00%
    Price
    $10 input / $50 output per 1M tokens
    Speed
    Up to 43.18 tok/s via Anthropic

    Strengths

    • Ranks #1 of 113 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SWE-Bench Verified, not total model capability
  2. 02
    AN
    Claude Mythos PreviewAnthropic
    Score
    93.90%

    Strengths

    • Ranks #2 of 113 compared models
    • 99th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SWE-Bench Verified, not total model capability
  3. 03
    AN
    Anthropic
    Score
    88.60%
    Price
    $5.0 input / $25 output per 1M tokens
    Speed
    Up to 42.00 tok/s via Vertex AI

    Strengths

    • Ranks #3 of 113 compared models
    • 98th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SWE-Bench Verified, not total model capability
  4. 04
    AN
    Anthropic
    Score
    87.60%
    Price
    $5.0 input / $25 output per 1M tokens
    Speed
    Up to 42.00 tok/s via Vertex AI

    Strengths

    • Ranks #4 of 113 compared models
    • 97th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SWE-Bench Verified, not total model capability
  5. 05
    AN
    Anthropic
    Score
    85.20%
    Price
    $2.0 input / $10 output per 1M tokens
    Speed
    Up to 42.00 tok/s via Anthropic

    Strengths

    • Ranks #5 of 113 compared models
    • 96th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SWE-Bench Verified, not total model capability

Selection summary

Best AI Models for SWE-Bench Verified

Claude Fable 5 currently leads SWE-Bench Verified with 95.00%. 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 Opus 4.8
Claude Opus 4.7
Claude Sonnet 5
Benchmark rank #1Claude Fable 595.00% · $10 input / $50 output per 1M tokens
Benchmark rank #2Claude Mythos Preview93.90%
Benchmark rank #3Claude Opus 4.888.60% · $5.0 input / $25 output per 1M tokens