Unspecified benchmark
LM Arena Agent Tool Hallucination Leaderboard
LM Arena Agent Tool Hallucination is a current leaderboard reporting Agent Score as a ratio.
Updated Oct 23, 2026
Models44
Model coverage44
MetricAgent Score
EvidenceA
LM Arena Agent Tool Hallucination Ranking
Higher agent score ranks better on this benchmark.
Sort by
Rank | Model | Agent Score | Percentile | Participants | Observations | Evidence | Evaluated |
|---|
| Rank01 | ModelMA | Agent Score0.71% (0.65% - 0.77%) | Percentile100.00% | Participants44 | Observations12.1M | EvidenceA | Evaluated |
| Rank02 | ModelZA | Agent Score0.69% (0.63% - 0.75%) | Percentile97.67% | Participants44 | Observations6.7M | EvidenceA | Evaluated |
| Rank03 | ModelOP | Agent Score0.67% (0.60% - 0.75%) | Percentile95.35% | Participants44 | Observations1M | EvidenceA | Evaluated |
| Rank04 | ModelAN | Agent Score0.67% (0.59% - 0.75%) | Percentile93.02% | Participants44 | Observations2.5M | EvidenceA | Evaluated |
| Rank05 | ModelAN | Agent Score0.66% (0.58% - 0.74%) | Percentile90.70% | Participants44 | Observations3.1M | EvidenceA | Evaluated |
| Rank06 | ModelZA | Agent Score0.66% (0.58% - 0.74%) | Percentile88.37% | Participants44 | Observations8.9M | EvidenceA | Evaluated |
| Rank07 | ModelZA | Agent Score0.65% (0.59% - 0.72%) | Percentile86.05% | Participants44 | Observations10.6M | EvidenceA | Evaluated |
| Rank08 | ModelXA | Agent Score0.64% (0.57% - 0.70%) | Percentile83.72% | Participants44 | Observations2.5M | EvidenceA | Evaluated |
| Rank09 | ModelOP | Agent Score0.62% (0.48% - 0.76%) | Percentile81.40% | Participants44 | Observations713.9K | EvidenceA | Evaluated |
| Rank10 | ModelME | Agent Score0.61% (0.54% - 0.69%) | Percentile79.07% | Participants44 | Observations6.9M | EvidenceA | Evaluated |
| Rank11 | ModelAN | Agent Score0.61% (0.48% - 0.73%) | Percentile76.74% | Participants44 | Observations1.9M | EvidenceA | Evaluated |
| Rank12 | ModelOP | Agent Score0.57% (0.48% - 0.67%) | Percentile74.42% | Participants44 | Observations4.4M | EvidenceA | Evaluated |
| Rank13 | ModelXA | Agent Score0.56% (0.47% - 0.66%) | Percentile72.09% | Participants44 | Observations3.3M | EvidenceA | Evaluated |
| Rank14 | ModelGO | Agent Score0.56% (0.47% - 0.65%) | Percentile69.77% | Participants44 | Observations5M | EvidenceA | Evaluated |
| Rank15 | ModelOP | Agent Score0.56% (0.41% - 0.71%) | Percentile67.44% | Participants44 | Observations2.6M | EvidenceA | Evaluated |
| Rank16 | ModelDE | Agent Score0.55% (0.42% - 0.68%) | Percentile65.12% | Participants44 | Observations765.7K | EvidenceA | Evaluated |
| Rank17 | ModelOP | Agent Score0.50% (0.36% - 0.64%) | Percentile62.79% | Participants44 | Observations1.6M | EvidenceA | Evaluated |
| Rank18 | ModelGO | Agent Score0.46% (0.32% - 0.60%) | Percentile60.47% | Participants44 | Observations4.7M | EvidenceA | Evaluated |
| Rank19 | ModelOP | Agent Score0.46% (0.28% - 0.63%) | Percentile58.14% | Participants44 | Observations3.8M | EvidenceA | Evaluated |
| Rank20 | ModelAN | Agent Score0.43% (0.28% - 0.58%) | Percentile55.81% | Participants44 | Observations3.5M | EvidenceA | Evaluated |
| Rank21 | ModelOP | Agent Score0.34% (0.10% - 0.58%) | Percentile53.49% | Participants44 | Observations3.8M | EvidenceA | Evaluated |
| Rank22 | ModelGO | Agent Score0.29% (0.14% - 0.44%) | Percentile51.16% | Participants44 | Observations3.4M | EvidenceA | Evaluated |
| Rank23 | ModelME | Agent Score0.21% (0.11% - 0.31%) | Percentile48.84% | Participants44 | Observations7.3M | EvidenceA | Evaluated |
| Rank24 | ModelME | Agent Score0.18% (-0.02% - 0.39%) | Percentile46.51% | Participants44 | Observations3M | EvidenceA | Evaluated |
| Rank25 | ModelTE | Agent Score0.12% (-0.06% - 0.29%) | Percentile44.19% | Participants44 | Observations8.3M | EvidenceA | Evaluated |
| Rank26 | ModelGO | Agent Score-0.10% (-0.75% - 0.55%) | Percentile41.86% | Participants44 | Observations2.7M | EvidenceA | Evaluated |
| Rank27 | ModelXI | Agent Score-0.14% (-0.39% - 0.11%) | Percentile39.53% | Participants44 | Observations3.2M | EvidenceA | Evaluated |
| Rank28 | ModelAC | Agent Score-0.14% (-0.50% - 0.21%) | Percentile37.21% | Participants44 | Observations2M | EvidenceA | Evaluated |
| Rank29 | ModelAC | Agent Score-0.15% (-0.35% - 0.05%) | Percentile34.88% | Participants44 | Observations7.1M | EvidenceA | Evaluated |
| Rank30 | ModelAC | Agent Score-0.70% (-0.84% - -0.56%) | Percentile32.56% | Participants44 | Observations13.5M | EvidenceA | Evaluated |
LM Arena Agent Tool Hallucination Highlights
The leading models and scores on this benchmark.
LM Arena Agent Tool Hallucination Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for LM Arena Agent Tool Hallucination
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is LM Arena Agent Tool Hallucination?
What LM Arena Agent Tool Hallucination measures and how its scores work.
A current leaderboard for LM Arena Agent Tool Hallucination.
It reports Agent Score, measured as a ratio.
Scores are shown in ratio. This benchmark is verified and has an evidence level of A.
- Family
- lm-arena-agent-tool-hallucination
- Modality
- Unspecified
- Primary category
- Unspecified
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LM Arena. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about LM Arena Agent Tool Hallucination.
Which model scores highest on LM Arena Agent Tool Hallucination?
Kimi K3 is currently ranked first with 0.71%.
What are the top three models on LM Arena Agent Tool Hallucination?
The current leaders are Kimi K3 (0.71%), GLM-5.2 (0.69%), and GPT-6 Astra (0.67%).
Which LM Arena Agent Tool Hallucination model has the lowest official input price?
GPT-6 Luna has the lowest matched official input price at $0.10 input / $0.50 output per 1M tokens.
Which models are fastest among LM Arena Agent Tool Hallucination results?
The fastest matched records are Gemini 3.7 Flash (409.96 tok/s via Google), GPT-5.6 Luna (215.52 tok/s via OpenAI), and Gemini 3.1 Pro (90.00 tok/s via Google).
Does the highest agent 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.