reasoning benchmark
MCP Atlas Leaderboard
MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities across complex multi-step tasks.
Updated Sep 7, 2026
Models34
Model coverage34
MetricScore
EvidenceB
MCP Atlas Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelME | Score88.10% | Percentile100.00% | Participants34 | EvidenceC | Evaluated |
| Rank02 | ModelMA | Score84.20% | Percentile96.97% | Participants34 | EvidenceC | Evaluated |
| Rank03 | ModelBY | Score83.80% | Percentile93.94% | Participants34 | EvidenceC | Evaluated |
| Rank04 | ModelTE | Score83.70% | Percentile90.91% | Participants34 | EvidenceC | Evaluated |
| Rank05 | ModelGO | Score83.60% | Percentile87.88% | Participants34 | EvidenceC | Evaluated |
| Rank06 | ModelAN | Score82.20% | Percentile84.85% | Participants34 | EvidenceC | Evaluated |
| Rank07 | ModelBY | Score80.30% | Percentile81.82% | Participants34 | EvidenceC | Evaluated |
| Rank08 | ModelTM | Score79.60% | Percentile78.79% | Participants34 | EvidenceC | Evaluated |
| Rank09 | ModelTE | Score79.10% | Percentile75.76% | Participants34 | EvidenceC | Evaluated |
| Rank10 | ModelAN | Score77.30% | Percentile72.73% | Participants34 | EvidenceC | Evaluated |
| Rank11 | ModelZA | Score76.80% | Percentile69.70% | Participants34 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score76.40% | Percentile66.67% | Participants34 | EvidenceC | Evaluated |
| Rank13 | ModelMA | Score76.00% | Percentile63.64% | Participants34 | EvidenceC | Evaluated |
| Rank14 | ModelME | Score75.50% | Percentile60.61% | Participants34 | EvidenceC | Evaluated |
| Rank15 | ModelOP | Score75.30% | Percentile57.58% | Participants34 | EvidenceC | Evaluated |
| Rank16 | ModelMI | Score74.20% | Percentile54.55% | Participants34 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score74.10% | Percentile51.52% | Participants34 | EvidenceC | Evaluated |
| Rank18 | ModelDE | Score73.60% | Percentile48.48% | Participants34 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score73.20% | Percentile45.45% | Participants34 | EvidenceC | Evaluated |
| Rank20 | ModelZA | Score71.80% | Percentile42.42% | Participants34 | EvidenceC | Evaluated |
| Rank21 | ModelGO | Score69.20% | Percentile39.39% | Participants34 | EvidenceC | Evaluated |
| Rank22 | ModelDE | Score69.00% | Percentile36.36% | Participants34 | EvidenceC | Evaluated |
| Rank23 | ModelZA | Score67.80% | Percentile33.33% | Participants34 | EvidenceC | Evaluated |
| Rank24 | ModelDE | Score67.40% | Percentile30.30% | Participants34 | EvidenceC | Evaluated |
| Rank25 | ModelOP | Score67.20% | Percentile27.27% | Participants34 | EvidenceC | Evaluated |
| Rank26 | ModelAC | Score62.80% | Percentile24.24% | Participants34 | EvidenceC | Evaluated |
| Rank27 | ModelAN | Score62.70% | Percentile21.21% | Participants34 | EvidenceC | Evaluated |
| Rank28 | ModelAN | Score62.30% | Percentile18.18% | Participants34 | EvidenceC | Evaluated |
| Rank29 | ModelAN | Score61.30% | Percentile15.15% | Participants34 | EvidenceC | Evaluated |
| Rank30 | ModelOP | Score60.60% | Percentile12.12% | Participants34 | EvidenceC | Evaluated |
MCP Atlas Highlights
The leading models and scores on this benchmark.
MCP Atlas Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MCP Atlas
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MCP Atlas?
What MCP Atlas measures and how its scores work.
MCP Atlas is a reasoning benchmark for evaluating AI models on scaled tool use capabilities.
It measures how well models coordinate and utilize multiple tools across complex multi-step tasks.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MCP Atlas
- 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 MCP Atlas.
Which model scores highest on MCP Atlas?
Muse Spark 1.1 is currently ranked first with 88.10%.
What are the top three models on MCP Atlas?
The current leaders are Muse Spark 1.1 (88.10%), Kimi K3 (84.20%), and Seed 2.1 Pro (83.80%).
Which MCP Atlas model has the lowest official input price?
DeepSeek-V4-Flash-Max has the lowest matched official input price at $0.14 input / $0.28 output per 1M tokens.
Which models are fastest among MCP Atlas 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.