THUDM/AgentBench
A Comprehensive Benchmark to Evaluate LLMs as Agents (ICLR'24)
View on GitHub →PythonApache-2.0chatgptgpt-4llmllm-agent
Stars
3.6k
Δ 7d
+25
Δ 30d
+80
Age
2 yrs
Last push
158d ago
90-day star history
Most-wanted open issues
Ranked by 👍 reactions. 62 open issues total, checked 2026-07-17.
| Issue | 👍 | Comments | Opened |
|---|---|---|---|
Benchmark for mistral models enhancement | 2 | 1 | 2 yrs |
| 2 | 1 | 1 yr | |
cg和kg都遇到了Worker not responding bug, help wanted | 1 | 1 | 2 yrs |
[Bug/Assistance] A lot of os-std tasks are impossible bug, help wanted | 1 | 1 | 1 yr |
| 1 | 0 | 1 yr |
Alternatives in Evals & benchmarks
非线智能 NoneLinear - ReLE评测：中文AI大模型能力评测（持续更新）：目前已囊括374个大模型，覆盖chatgpt、gpt-5.4、谷歌gemini-3.1-pro、Claude-4.6、文心ERNIE-…
⭐6.3k
+327d
3 yrs
Measuring frontier coding agents on original, long-horizon engineering tasks
⭐1.2k
+837d
Python
2 mo
Measuring how well CLI agents like Claude Code or Codex CLI can post-train base LLMs on a single H100 GPU in 1…
⭐455
+247d
Python
7 mo
GateMem: a benchmark and evaluation toolkit for memory governance in multi-principal shared-memory LLM agents.
⭐191
+277d
Python
1 mo
Pier is a Harbor fork built for DeepSWE, with stronger support for CLI agents in air-gapped (no-internet) task…
⭐116
+137d
Python
2 mo