Alternatives to AI21 Labs

Google's developer gateway to its large language models, launched March 2023 with the PaLM API and MakerSuite prototyping tool.

Mistral sells sovereign AI: frontier open-weight models you can own, self-host, or run on EU inference.

Zhipu AI builds the open-source GLM model family and a full-stack MaaS platform for coding, multimodal, and long-horizon agent work.

Apache 2.0 open-weight model family from TII Abu Dhabi, spanning hybrid Transformer-Mamba, Arabic, reasoning, and multimodal vision models.

Chinese medical LLM vendor with open commercial weights from 7B to 235B and the 百小医 AI family-doctor app.

Korean enterprise LLM — consumer CLOVA X service ended April 2026

Liquid AI's open-weight LFM2.5 model family runs native text, vision, and audio AI locally on CPU, GPU, or NPU.

Baichuan-7B is a free 7B bilingual Chinese-English base LLM released in 2023 on Hugging Face for fine-tuning and benchmark research.

Databricks' open-source instruction-following LLM, built on Pythia-12B and fine-tunable on a single A100 in about 30 minutes.

StableLM: Stability AI's open-weight language model suite from 2023, self-hosted and licensed for commercial or research use

Deploy and discover AI agents on an open marketplace built for SEO-driven reach across multiple AI assistants.

Open-weight 35B Mixture-of-Experts model that activates about 3B parameters per token, so agentic coding and reasoning can run on local hardware.

MiniMax M3 packs a 1M-token context, native multimodality, and frontier coding into a token subscription.

Open-source 27B model for agentic coding and multimodal reasoning, self-hosted under Apache 2.0.

Cohere delivers enterprise AI — Command generation, Embed 5 retrieval, and Rerank — deployed inside your own VPC or on-prem.

Qwen's early-access flagship preview model for agentic coding, tool use, and precise instruction following.

Open-weight multimodal LLMs — HPT 1.5 Edge (~4B) and HPT 1.5 Air (8B on Llama 3) — that you download and self-host.

DeepSeek is a free reasoning and web-search chat built on the V4.1-Flash multimodal model, plus a usage-billed developer API.