Alternatives to Zhipu GLM

Zhipu AI's GLM-5.3 coding model and MaaS platform deliver open-source SOTA front-end programming, 1M context, and autonomous agents.

Falcon LLM delivers Apache-2.0 open-weight language and multimodal models you can self-host, from hybrid Transformer-Mamba to Arabic ASR

01.AI Yi: Chinese enterprise AI decision platforms and the Yi model stack behind them

Open-source multimodal function calling that lets agents see images and act on them in one pass.

LFM2.5 is Liquid AI's open-weight foundation model family for on-device AI — native text, vision, and audio inference on CPU, GPU, or NPU.

StableLM: Stability AI's open-weight language model for self-hosted text and code generation

PaLM API gives Google Cloud developers safe, scalable access to Google's language models, with MakerSuite for prompt prototyping.

Poolside builds open-weight agentic coding models — Laguna XS 2.1 (33B) and Laguna S 2.1 (118B) — for code that stays inside your security

Mistral is the sovereign AI stack: frontier open-weight models you can self-host, fine-tune, or run in-region.

Open-weight agentic coding models — Laguna XS 2.1 at 33B on-device and Laguna S 2.1 at 118B for long-horizon agent work.

Open-weight coding model for 12-hour autonomous agent runs and 300-agent swarm orchestration.

MIT-licensed AI agent that runs its full tool loop, sandboxed Linux VM, and cron scheduler inside a single browser tab.

Open-source MCP server that gives coding agents live Chrome DevTools for debugging, traces, and browser automation.

StepFun's OpenAI-compatible API bundles frontier reasoning, vision-language parsing and a full speech stack — realtime voice, ASR, TTS and

Together Compute is an OpenAI-compatible inference API for serving open-weight AI models at published per-token rates.

Open-weight 35B Mixture-of-Experts model with roughly 3B active parameters per token, built for local agentic coding and reasoning.

Devstral 2 is Mistral's open-weight coding model family — 123B and 24B sizes plus the Apache 2.0 Mistral Vibe CLI terminal agent

Alibaba's open-weight 27B model for agentic coding and image-plus-text reasoning, self-hosted under Apache 2.0.

Open-source agentic AI framework that runs autonomous agents with real terminal, file, and browser access on your own machine.

RobBERT is a family of open pre-trained Dutch language models you fine-tune for classification, regression, and token-tagging tasks.

An open-source autonomous AI companion that learns your habits and drives your desktop, browser and APIs for you.

Open-source Python library for loading, fine-tuning, and running transformer models across text, vision, audio, and video.

SmallDoge is a community open-source project publishing compact small language models, training checkpoints, and curated datasets for fast local inference.

Paper Instruments (formerly The LLM Data Company) ships open-weight frontier models for domain knowledge work, trained with on-policy reinforcement learning.

Reka builds omni models and edge infrastructure for real-time video reasoning on hardware you already own.

Aleph Alpha Pharia delivers sovereign specialised LLMs, including the open-weight Kolibri model, on EU-hosted infrastructure.

OpenAI Agents SDK: a free, MIT-licensed Python framework for building multi-agent workflows with handoffs, guardrails, and tracing.

Mastra is an open-source TypeScript agent framework for durable agents, typed workflows, and built-in evals.

Pipedream is the integration layer that gives AI agents managed auth plus 10,000+ tools across 3,000+ APIs.

Tessl is an agent enablement platform for versioning, scanning and evaluating the AI agent skills your coding agents run.