OpenCodePapers

agents-on-toolathlon-verified-108-tasks

Agents
Dataset Link
Results over time
Click legend items to toggle metrics. Hover points for model names.
Leaderboard
Click a sortable column header to sort.
PaperCodePass@1ModelNameReleaseDate
Introducing Claude Opus 5✓ Link0.806Claude Opus 52026-07-24
GLM-5.3-Flash: Frontier Intelligence, Flash Cost✓ Link0.784GLM-5.3-Flash2026-08-26
Introducing Claude Opus 5.5✓ Link0.778Claude Opus 5.52026-09-22
Introducing Claude Sonnet 5.5✓ Link0.778Claude Sonnet 5.52026-09-28
MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement✓ Link0.769MiMo-V2.6-Pro-RL2026-10-08
Kimi K3: Open Frontier Intelligence✓ Link0.765Kimi K3 Max2026-07-27
Introducing Muse Spark 1.1✓ Link0.756Muse Spark 1.12026-07-09
Qwen3.8-Flash-Next✓ Link0.735Qwen3.8-Flash-Next2026-08-24
GLM-5.3: Frontier Coding with Emergent Cyber Capabilities✓ Link0.73GLM-5.32026-08-14
Introducing K2 Horizon: Frontier Performance, Radically Open✓ Link0.653K2-Horizon-375B-A23B2026-09-03
Inkling-Small✓ Link0.544Inkling-Small2026-07-27
WEFT: Scaling Tool-Use Post-Training for General-Purpose Agents0.4599WEFT-35B-A3B2026-09-29
ProgAgent: Learning Tool Orchestration through Programmatic Training✓ Link0.171ProgAgent-14B2026-10-10
Check out how to contribute new results. Then edit this file on GitLab or GitHub.