# ToolSandbox leaderboard

Agentic benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelchoice.dev/leaderboards/benchmarks/agentic-tool-sandbox

Every model the catalog carries a published ToolSandbox value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Agentic |
| Measure | Average accuracy |
| Tasks | Stateful tool-use conversations |
| Difficulty | Tool use with changing state |
| Published by | [ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilities](https://github.com/apple-aiml-research/ToolSandbox) |

Apple benchmark of stateful, conversational tool use, where tool calls change state and a simulated user takes part.

## ToolSandbox ranking

3 models with a published ToolSandbox value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Average accuracy | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [LFM3-1.7B](https://modelchoice.dev/models/lfm3-1-7b) | LiquidAI | 80.36 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Nemotron 3.5 Bolt E4B](https://modelchoice.dev/models/nemotron-3-5-bolt-e4b) | NVIDIA | 74.13 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Nemotron 3.5 Bolt E4B BF16](https://modelchoice.dev/models/nemotron-3-5-bolt-e4b-bf16) | NVIDIA | 68.77 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
