Gaia2 evaluates assistants coordinating tasks across simulated email, calendars, shopping, contacts, and other apps. It tests execution, search, adaptation, ambiguity handling, and time constraints. This ranking uses the Gaia2-CLI leaderboard with the OpenClaw 2026.4.1 harness and reports overall pass@1. The older vanilla, noise, and agent-to-agent results are separate evaluations.
| Model / creator | Version | Reported score |
|---|---|---|
 Claude Opus 4.6Anthropic · Proprietary | CLI · OpenClaw 2026.4.1 | 57% |
 GPT-5.5OpenAI · Proprietary | CLI · OpenClaw 2026.4.1 | 56.4% |
| GPT-5.4OpenAI · Proprietary | CLI · OpenClaw 2026.4.1 | 55.6% |
 Gemini 3.1 ProGoogle · Proprietary | CLI · OpenClaw 2026.4.1 | 52% |
| Claude Sonnet 4.6Anthropic · Proprietary | CLI · OpenClaw 2026.4.1 | 51.9% |
 GLM-5.1Z.ai · Open source / open weights | CLI · OpenClaw 2026.4.1 | 50.5% |
 Kimi K2.5Moonshot AI · Open source / open weights | CLI · OpenClaw 2026.4.1 | 34% |
api host: openai · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: xhigh · reasoning context: Not specified
Meta Agents Research Environments ↗api host: google · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: high · reasoning context: Not specified
Meta Agents Research Environments ↗api host: OpenRouter · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: enabled · reasoning context: Not round-tripped by this harness for OpenRouter; may affect multi-step performance
Meta Agents Research Environments ↗api host: anthropic · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: high · reasoning context: Not specified
Meta Agents Research Environments ↗api host: openai · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: high · reasoning context: Not specified
Meta Agents Research Environments ↗api host: anthropic · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: high · reasoning context: Not specified
Meta Agents Research Environments ↗api host: OpenRouter · capabilities: search, execution, adaptability, ambiguity, time · edition: Gaia2-CLI · harness: OpenClaw 2026.4.1 · reasoning: enabled · reasoning context: Not round-tripped by this harness for OpenRouter; may affect multi-step performance
Meta Agents Research Environments ↗LiveBench 2.0 evaluates data analysis on its October 25, 2026 release.
LifeValLifeVal evaluates personal assistants on everyday responsibilities, including caregiving, household administration, shopping, education, personal finances, and accessing government services.
LiveBench: ResearchLiveBench 2.0 evaluates research reports on its October 25, 2026 release.
