reasoning benchmark
Toolathlon Leaderboard
Toolathlon is a benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories.
Updated Sep 5, 2026
Models41
Model coverage41
MetricScore
EvidenceB
Toolathlon Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelZA | Score78.40% | Percentile100.00% | Participants41 | EvidenceC | Evaluated |
| Rank02 | ModelME | Score75.60% | Percentile97.50% | Participants41 | EvidenceC | Evaluated |
| Rank03 | ModelDE | Score74.10% | Percentile95.00% | Participants41 | EvidenceC | Evaluated |
| Rank04 | ModelTE | Score74.10% | Percentile92.50% | Participants41 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score73.50% | Percentile90.00% | Participants41 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score73.50% | Percentile87.50% | Participants41 | EvidenceC | Evaluated |
| Rank07 | ModelMA | Score73.20% | Percentile85.00% | Participants41 | EvidenceC | Evaluated |
| Rank08 | ModelZA | Score73.00% | Percentile82.50% | Participants41 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score72.50% | Percentile80.00% | Participants41 | EvidenceC | Evaluated |
| Rank10 | ModelDE | Score70.30% | Percentile77.50% | Participants41 | EvidenceC | Evaluated |
| Rank11 | ModelAN | Score59.90% | Percentile75.00% | Participants41 | EvidenceC | Evaluated |
| Rank12 | ModelOP | Score58.00% | Percentile72.50% | Participants41 | EvidenceC | Evaluated |
| Rank13 | ModelGO | Score56.50% | Percentile70.00% | Participants41 | EvidenceC | Evaluated |
| Rank14 | ModelOP | Score55.60% | Percentile67.50% | Participants41 | EvidenceC | Evaluated |
| Rank15 | ModelOP | Score54.60% | Percentile65.00% | Participants41 | EvidenceC | Evaluated |
| Rank16 | ModelTM | Score54.40% | Percentile62.50% | Participants41 | EvidenceC | Evaluated |
| Rank17 | ModelAN | Score54.30% | Percentile60.00% | Participants41 | EvidenceC | Evaluated |
| Rank18 | ModelOP | Score53.40% | Percentile57.50% | Participants41 | EvidenceC | Evaluated |
| Rank19 | ModelOP | Score53.10% | Percentile55.00% | Participants41 | EvidenceC | Evaluated |
| Rank20 | ModelDE | Score51.80% | Percentile52.50% | Participants41 | EvidenceC | Evaluated |
| Rank21 | ModelBY | Score50.60% | Percentile50.00% | Participants41 | EvidenceC | Evaluated |
| Rank22 | ModelMA | Score50.00% | Percentile47.50% | Participants41 | EvidenceC | Evaluated |
| Rank23 | ModelPO | Score49.70% | Percentile45.00% | Participants41 | EvidenceC | Evaluated |
| Rank24 | ModelGO | Score49.40% | Percentile42.50% | Participants41 | EvidenceC | Evaluated |
| Rank25 | ModelBY | Score49.10% | Percentile40.00% | Participants41 | EvidenceC | Evaluated |
| Rank26 | ModelTE | Score48.50% | Percentile37.50% | Participants41 | EvidenceC | Evaluated |
| Rank27 | ModelZA | Score48.20% | Percentile35.00% | Participants41 | EvidenceC | Evaluated |
| Rank28 | ModelDE | Score47.80% | Percentile32.50% | Participants41 | EvidenceC | Evaluated |
| Rank29 | ModelOP | Score46.30% | Percentile30.00% | Participants41 | EvidenceC | Evaluated |
| Rank30 | ModelMI | Score46.30% | Percentile27.50% | Participants41 | EvidenceC | Evaluated |
Toolathlon Highlights
The leading models and scores on this benchmark.
Toolathlon Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for Toolathlon
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is Toolathlon?
What Toolathlon measures and how its scores work.
Toolathlon evaluates AI agents across ten different tool-use scenarios.
It measures proficiency in tool selection, sequencing, and execution, reported as a Score ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- Toolathlon
- 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 Toolathlon.
Which model scores highest on Toolathlon?
GLM-5.3-Flash is currently ranked first with 78.40%.
What are the top three models on Toolathlon?
The current leaders are GLM-5.3-Flash (78.40%), Muse Spark 1.1 (75.60%), and DeepSeek-V4-Pro-0813 (74.10%).
Which Toolathlon model has the lowest official input price?
GLM-5.3-Flash has the lowest matched official input price at $0.08 input / $0.25 output per 1M tokens.
Which models are fastest among Toolathlon results?
The fastest matched records are GLM-5.3 (472.69 tok/s via FriendliAI), Gemini 3.5 Flash (210.94 tok/s via Google), and GPT-5.5 (134.94 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.