Skip to content
SPECIFICTION
Get listed

AgentBench vs LM Evaluation Harness

AgentBench

github.com

LM Evaluation Harness

github.com

Side by side

FigureAgentBenchLM Evaluation Harness
Price from—Free
Free plan—Yes
Free trial——
Open source——
LinuxYesYes
APIYesYes
Self-hostedYesYes
Plans published01
Spec score · LLM Evaluation Tools7.8 (#10)7.5 (#17)
PurposeAgentBench is a benchmark for evaluating the reasoning and decision-making abilities of LLMs acting as agents.LM Evaluation Harness is a unified framework for evaluating generative language models across evaluation tasks.
SupportThe project invites users to join its Slack for questions and collaboration, and lists a Google Group for questions about AgentBench FC.The project directs users to open a GitHub issue or join the EleutherAI Discord for support.

Plans

AgentBench

No plans published.

AgentBench pricing →

LM Evaluation Harness

  • LM Evaluation HarnessFree

LM Evaluation Harness pricing →

In their own words

AgentBench

AgentBench has no write-up on Specifiction yet; its figures are above.

Visit github.com

LM Evaluation Harness

LM Evaluation Harness is a free, open-source framework for evaluating generative language models across many tasks. It lists more than 60 standard academic benchmarks, with hundreds of subtasks and variants. Supported model options include Hugging Face Transformers, GPT-NeoX, Megatron-DeepSpeed, and vLLM. The harness also connects to hosted OpenAI and Anthropic APIs and to local servers compatible with OpenAI APIs. Users can define prompts and metrics, configure tasks with YAML, and extend model backends, filters, metrics, and aggregations through external packages or local modules. Publicly available prompts are intended to support reproducibility and comparison between papers. The base package contains the core framework; model backends are separate optional installs. It runs self-hosted on Linux and API environments, with documented Windows ML and ONNX Runtime accelerator options. MIT licensing applies. APIs without logits or log probabilities are limited to generate-until tasks.

Visit github.com