SlackBench - MCP Low - Summarization: leaderboard
Agents answer questions across realistic Slack workplace reenactments and associated evolving code repositories. Mean category pass score (%). SlackBench140tasks:58AgenticSearch,48Hallucination,14Summarization,20CircumstanceInference; three runs,20min/query; GPT-5.4 judge. Summarization earns checklist partial credit, other categories binary pass; Overall is the unweighted category mean. Harness, MCP/CLI access and reasoning effort remain separate in row identities. Native first-party model APIs.
Metric: Mean category pass score (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 (Benchmark Sweep Run: source=SlackBench; harness=Codex CLI; access=MCP; effort=low) | 51.52 | |
| 2 | Claude Opus 4.7 (Benchmark Sweep Run: source=SlackBench; harness=Claude Code; access=MCP; effort=low) | 45.19 | |
| 3 | Claude Sonnet 4.6 (Benchmark Sweep Run: source=SlackBench; harness=Claude Code; access=MCP; effort=low) | 40.86 | |
| 4 | Gemini 3.1 Pro (Preview) (Benchmark Sweep Run: source=SlackBench; harness=Gemini CLI; access=MCP; effort=low) | 27.04 | |
| 5 | GPT-5.4 Mini (Benchmark Sweep Run: source=SlackBench; harness=Codex CLI; access=MCP; effort=low) | 24.8 |
Interactive version: theaggregate.ai/benchmark?slug=slackbench-mcp-low-summarization · How It Works · Data refreshed daily, snapshot 2026-11-06.