CLM

CLM: Language model connects states and actions.

CLM is a language model that connects states and actions with a contrastive learning objective. It is trained on a large-scale dataset and serves as a zero-shot action classifier. CLM scores each action by how well its embedding aligns with the state embedding and selects the highest-scoring action. It is pre-trained on internet-scale Q&A, mid-trained on synthetic hard negatives, and post-trained on agentic traces. CLM can be easily fine-tuned on downstream tasks.

CLM at a glance

Best for: Software engineer, Researcher
Runs on: Web, CLI
Pricing: Free
Released: September 2026
Category: AI research tool
Website: github.com/contrastive-lm/clm

Key features

State encoder
Action encoder
Contrastive learning objective
Zero-shot action classification
Fine-tuning capability

Use it for

Decision-making
Action classification
Verification

Alternatives to CLM

Clef — Offers low latency and high-speed classification for agentic workflows.
LLM CLI — It enables command-line interaction with remote APIs and local machine models.
LLM Stats — Independent rankings updated continuously from live API metrics.
Overview — Learn language models by running hands-on experiments without writing scripts
LLM Wiki Research — Compiles parallel agent research into source-backed markdown knowledge bases.
Ncm — Generates text significantly faster than diffusion baselines while maintaining high quality.
OpenCompass — It offers an objective, comprehensive, and credible open-source AI evaluation system.
GEAR — Organizes recursive self-improvement through evolutionary search and reproducible evaluation.
Vegaml — Runs locally on your hardware with calibrated probabilities and conformal answer sets.
Homi — It answers complex audio questions natively with high accuracy.
Deepopen — Provides millisecond-level inference speeds and completely eliminates text generation hallucinations.
ContextGem 1.0 — It extracts structured data with source tracing and uses both LLMs and decision models.

All CLM alternatives