Arena ranking
Agent Tool Hallucination Leaderboard
Browse Agent Tool Hallucination results, model ratings and supporting selection context.
Data as of 2026-09-08
Agent Tool Hallucination Ranking
This source leaderboard ranking orders models by their original Arena rank rather than an aggregated benchmark score.
Sort by
Rank | Model | Source Agent score | LLMBoard Score | Observations | Official input / 1M | Output speed |
|---|
| Rank01 | ModelOP | Source Agent score0.78% | LLMBoard ScoreN/A | Observations3M | Official input / 1M$4 | Output speed2.36 tok/s |
| Rank02 | ModelOP | Source Agent score0.78% | LLMBoard ScoreN/A | Observations981.2K | Official input / 1M$2 | Output speed99.35 tok/s |
| Rank03 | ModelXA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.8M | Official input / 1M$2 | Output speed132.69 tok/s |
| Rank04 | ModelDE | Source Agent score0.78% | LLMBoard ScoreN/A | Observations2.7M | Official input / 1MN/A | Output speed45.48 tok/s |
| Rank05 | ModelOP | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.5M | Official input / 1M$0.20 | Output speed41.87 tok/s |
| Rank06 | ModelXA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.6M | Official input / 1M$2 | Output speed80.00 tok/s |
| Rank07 | ModelOP | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.8M | Official input / 1M$5 | Output speed134.94 tok/s |
| Rank08 | ModelMA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations7.8M | Official input / 1M$3 | Output speedN/A |
| Rank09 | ModelZA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations3.7M | Official input / 1M$1.4 | Output speed3.96 tok/s |
| Rank10 | ModelZA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.8M | Official input / 1M$0.075 | Output speed74.28 tok/s |
| Rank11 | ModelMA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations304.3K | Official input / 1M$0.95 | Output speed1.24 tok/s |
| Rank12 | ModelMA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations383.4K | Official input / 1M$0.95 | Output speedN/A |
| Rank13 | ModelAN | Source Agent score0.78% | LLMBoard ScoreN/A | Observations488.8K | Official input / 1M$10 | Output speed7.67 tok/s |
| Rank14 | ModelZA | Source Agent score0.78% | LLMBoard ScoreN/A | Observations4M | Official input / 1M$1.4 | Output speed472.69 tok/s |
| Rank15 | ModelOP | Source Agent score0.78% | LLMBoard ScoreN/A | Observations3.3M | Official input / 1M$2.5 | Output speed50.00 tok/s |
| Rank16 | ModelAN | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.2M | Official input / 1M$5 | Output speed123.15 tok/s |
| Rank17 | ModelAN | Source Agent score0.78% | LLMBoard ScoreN/A | Observations1.7M | Official input / 1M$10 | Output speed43.18 tok/s |
| Rank18 | ModelME | Source Agent score0.77% | LLMBoard ScoreN/A | Observations798.1K | Official input / 1M$1.25 | Output speed28.44 tok/s |
| Rank19 | ModelME | Source Agent score0.77% | LLMBoard ScoreN/A | Observations3.2M | Official input / 1M$1.25 | Output speed14.00 tok/s |
| Rank20 | ModelAN | Source Agent score0.76% | LLMBoard ScoreN/A | Observations2.2M | Official input / 1M$5 | Output speed62.88 tok/s |
| Rank21 | ModelDE | Source Agent score0.76% | LLMBoard ScoreN/A | Observations5.4M | Official input / 1M$0.14 | Output speed15.91 tok/s |
| Rank22 | ModelAN | Source Agent score0.72% | LLMBoard ScoreN/A | Observations1.4M | Official input / 1M$5 | Output speed42.00 tok/s |
| Rank23 | ModelGO | Source Agent score0.71% | LLMBoard ScoreN/A | Observations1.6M | Official input / 1M$0.75 | Output speed117.86 tok/s |
| Rank24 | ModelAN | Source Agent score0.71% | LLMBoard ScoreN/A | Observations1.3M | Official input / 1M$3 | Output speed7.67 tok/s |
| Rank25 | ModelGO | Source Agent score0.70% | LLMBoard ScoreN/A | Observations531.5K | Official input / 1M$0.75 | Output speed51.56 tok/s |
| Rank26 | ModelGO | Source Agent score0.69% | LLMBoard ScoreN/A | Observations922K | Official input / 1M$0.75 | Output speed62.56 tok/s |
| Rank27 | ModelXA | Source Agent score0.67% | LLMBoard ScoreN/A | Observations1.1M | Official input / 1M$1.25 | Output speedN/A |
| Rank28 | ModelTE | Source Agent score0.66% | LLMBoard ScoreN/A | Observations1.2M | Official input / 1MN/A | Output speedN/A |
| Rank29 | ModelAN | Source Agent score0.63% | LLMBoard ScoreN/A | Observations2.7M | Official input / 1M$2 | Output speed42.00 tok/s |
| Rank30 | ModelMI | Source Agent score0.61% | LLMBoard ScoreN/A | Observations772.3K | Official input / 1M$0.30 | Output speedN/A |
Source leaderboard overview
This leaderboard overview compares the source ranking, score gaps and observed task volume.
Ranked models49
SignalAgent score
Agent score and confidence
Use confidence and observed task volume to judge how firmly each leaderboard position is supported; these source proportions remain separate from benchmark coverage.
0.78%
0.78%
0.78%
0.78%
0.78%
0.78%
0.78%
0.78%
0.78%
Agent Tool Hallucination Leaderboard Vendor Mix
Vendor representation among the top-ranked models in this leaderboard.
Top-entry vendor mix
25 modelsThe Top AI Models for Agent Tool Hallucination
A practical summary of the first five entries in this agent tool hallucination AI model leaderboard, with benchmark evidence and pricing kept in context.
About this leaderboard ranking
Common questions about the Agent Tool Hallucination Leaderboard.
What is the top model on Agent Tool Hallucination Leaderboard?
GPT-5.6-Sol is currently ranked first with 0.78 Agent score.
What are the top three models on Agent Tool Hallucination Leaderboard?
The current leaders are GPT-5.6-Sol (rank #1), GPT-5.6-Terra (rank #2), and Grok 4.6 (rank #3).
Which ranked model has the lowest official input price?
GLM 5.3 Flash has the lowest matched official input price at $0.075 per 1M tokens.
Which models are fastest on this leaderboard?
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 top source result mean the model is best at every task?
No. Source Arena results describe a particular task or preference signal. Capability benchmarks, prices and runtime can produce different rankings.