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
1464 models816 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

agents benchmark

Ramp SWE-Bench v2 Leaderboard

Ramp SWE-Bench v2 is a private coding benchmark of 500 backend engineering tasks drawn from pull requests merged into production after engineer review.

Updated Nov 22, 2026

Models16
Model coverage16
MetricScore
EvidenceB

On this page

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

Ramp SWE-Bench v2 Ranking

Higher score ranks better on this benchmark.

16 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelOPGPT-6.2 AstraOpenAIScore76.40%Percentile100.00%Participants16EvidenceBEvaluatedNov 21, 2026
Rank02ModelGOGemini 4 CarbonGoogleScore75.40%Percentile93.33%Participants16EvidenceBEvaluatedNov 21, 2026
Rank03ModelANClaude Fable 5.2AnthropicScore73.40%Percentile86.67%Participants16EvidenceBEvaluatedNov 21, 2026
Rank04ModelOPGPT-6.2 SolOpenAIScore73.40%Percentile80.00%Participants16EvidenceBEvaluatedNov 21, 2026
Rank05ModelGOGemini 4 HydrogenGoogleScore71.60%Percentile73.33%Participants16EvidenceBEvaluatedNov 21, 2026
Rank06ModelOPGPT-6.1 TerraOpenAIScore70.40%Percentile66.67%Participants16EvidenceBEvaluatedNov 21, 2026
Rank07ModelANClaude Opus 5.5AnthropicScore67.80%Percentile60.00%Participants16EvidenceBEvaluatedNov 21, 2026
Rank08ModelXAGrok 4.7xAIScore64.40%Percentile53.33%Participants16EvidenceBEvaluatedNov 21, 2026
Rank09ModelOPGPT-6 AstraOpenAIScore63.40%Percentile46.67%Participants16EvidenceBEvaluatedNov 21, 2026
Rank10ModelANClaude Sonnet 5.5AnthropicScore63.20%Percentile40.00%Participants16EvidenceBEvaluatedNov 21, 2026
Rank11ModelDEDeepSeek-V4.1-Flash-1017DeepSeekScore60.80%Percentile33.33%Participants16EvidenceBEvaluatedNov 21, 2026
Rank12ModelOPGPT-6.1 SolOpenAIScore59.60%Percentile26.67%Participants16EvidenceBEvaluatedNov 21, 2026
Rank13ModelMEMuse WonderMetaScore58.60%Percentile20.00%Participants16EvidenceBEvaluatedNov 21, 2026
Rank14ModelANClaude Haiku 5.5AnthropicScore57.40%Percentile13.33%Participants16EvidenceBEvaluatedNov 21, 2026
Rank15ModelOPGPT-6 LunaOpenAIScore52.40%Percentile6.67%Participants16EvidenceBEvaluatedNov 21, 2026
Rank16ModelMEMuse EmberMetaScore51.80%Percentile0.00%Participants16EvidenceBEvaluatedNov 21, 2026

Ramp SWE-Bench v2 Highlights

The leading models and scores on this benchmark.

Ramp SWE-Bench v2 Score Distribution

A closer view of the leading scores on this benchmark.

Ramp SWE-Bench v2

The Top AI Models for Ramp SWE-Bench v2

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

What is Ramp SWE-Bench v2?

What Ramp SWE-Bench v2 measures and how its scores work.

Ramp SWE-Bench v2 uses one-shot prompts in a sandboxed copy of Ramp's codebase, where each model produces a review-ready patch using its lab's native agent harness, with web tools disabled and high reasoning effort.

It reports Score as pass@1, based on whether a panel of LLM judges finds no material difference between the candidate patch and the engineer-reviewed merged solution where the merged solution is better.

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

Family
Ramp SWE-Bench v2
Modality
text
Primary category
agents
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 Ramp SWE-Bench v2.

Which model scores highest on Ramp SWE-Bench v2?

GPT-6.2 Astra is currently ranked first with 76.40%.

What are the top three models on Ramp SWE-Bench v2?

The current leaders are GPT-6.2 Astra (76.40%), Gemini 4 Carbon (75.40%), and Claude Fable 5.2 (73.40%).

Which Ramp SWE-Bench v2 model has the lowest official input price?

Claude Haiku 5.5 has the lowest matched official input price at $0.10 input / $0.50 output per 1M tokens.

Which models are fastest among Ramp SWE-Bench v2 results?

The fastest matched records are Muse Ember (63.51 tok/s via Meta Model API), Claude Sonnet 5.5 (42.00 tok/s via Anthropic), and GPT-6.1 Terra (34.75 tok/s via OpenAI).

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 Ramp SWE-Bench v2 measure?

It reports Score as pass@1, based on whether a panel of LLM judges finds no material difference between the candidate patch and the engineer-reviewed merged solution where the merged solution is better.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

16 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-6.2 Astra76.40%
Rank #2Gemini 4 Carbon75.40%
Rank #3Claude Fable 5.273.40%
Rank #4GPT-6.2 Sol73.40%

Ranking basisThis ramp swe-bench v2 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-6.2 AstraOpenAI
    Score
    76.40%
    Price
    $8.0 input / $40 output per 1M tokens

    Strengths

    • Ranks #1 of 16 compared models
    • 100th percentile on this benchmark
    • B evidence result

    Considerations

    • This result measures Ramp SWE-Bench v2, not total model capability
  2. 02
    GO
    Gemini 4 CarbonGoogle
    Score
    75.40%
    Price
    $2.0 input / $10 output per 1M tokens
    Speed
    Up to 11.31 tok/s via Google

    Strengths

    • Ranks #2 of 16 compared models
    • 93th percentile on this benchmark
    • B evidence result

    Considerations

    • This result measures Ramp SWE-Bench v2, not total model capability
  3. 03
    AN
    Anthropic
    Score
    73.40%
    Price
    $8.0 input / $40 output per 1M tokens

    Strengths

    • Ranks #3 of 16 compared models
    • 87th percentile on this benchmark
    • B evidence result

    Considerations

    • This result measures Ramp SWE-Bench v2, not total model capability
  4. 04
    OP
    OpenAI
    Score
    73.40%
    Price
    $2.0 input / $10 output per 1M tokens

    Strengths

    • Ranks #4 of 16 compared models
    • 80th percentile on this benchmark
    • B evidence result

    Considerations

    • This result measures Ramp SWE-Bench v2, not total model capability
  5. 05
    GO
    Google
    Score
    71.60%
    Price
    $1.5 input / $7.5 output per 1M tokens

    Strengths

    • Ranks #5 of 16 compared models
    • 73th percentile on this benchmark
    • B evidence result

    Considerations

    • This result measures Ramp SWE-Bench v2, not total model capability

Selection summary

Best AI Models for Ramp SWE-Bench v2

GPT-6.2 Astra currently leads Ramp SWE-Bench v2 with 76.40%. 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 Fable 5.2
GPT-6.2 Sol
Gemini 4 Hydrogen
Benchmark rank #1GPT-6.2 Astra76.40% · $8.0 input / $40 output per 1M tokens
Benchmark rank #2Gemini 4 Carbon75.40% · $2.0 input / $10 output per 1M tokens
Benchmark rank #3Claude Fable 5.273.40% · $8.0 input / $40 output per 1M tokens