Alternatives to Keras Hub

Reka builds omni models for real-time video reasoning that run on-device, not just in the cloud.

Open-weight agentic coding models — Laguna XS 2.1 and Laguna S 2.1 — built for secure on-prem and air-gapped enterprise AI.

Zhipu GLM delivers open-source LLM models, MaaS APIs, and autonomous agents for Chinese enterprises and developers.

Reka AI builds omni models, edge inference, and video infrastructure for the physical AI era.

Sarvam AI is India's sovereign AI platform for Indic-language speech, text, and document models, deployable on cloud, VPC, or air-gapped infrastructure.

Sovereign specialized LLMs: Aleph Alpha Pharia trains custom domain models on EU infrastructure for regulated European organizations.

Cohere delivers sovereign enterprise AI models — Command, Embed, Rerank, and North — deployable in your VPC, on-prem, or in Model Vault.

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

StepFun ships frontier and open-weight multimodal models plus a full voice stack — Realtime, ASR, TTS and music — behind one OpenAI-compatible endpoint.

Hugging Face is the open model hub where you host, discover and deploy 2M+ models, 500k+ datasets and 1M+ Spaces.

Together Compute is an AI-native cloud for running open-source models — serverless inference, batch jobs, model shaping, and GPU clusters in one platform.

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

Frontier code models built for ultra-long-context software engineering and AI research automation.

Enterprise AI platform that cuts token cost for agent workloads by routing across models and tuning small open models to frontier quality.

DeepSeek is a free reasoning and search chat with the V4.1-Flash multimodal model and a usage-billed developer API.

Enterprise AI decision platform turning CEO mandates into measurable business outcomes.

Mistral sells frontier open-weight AI models, long-horizon Vibe agents, and document intelligence you can self-host or run on EU

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

Korean enterprise LLM — consumer CLOVA X service ended April 2026

01.AI Yi is a China-focused enterprise AI platform sold through CEO-led transformation engagements, not self-serve API access.

Poolside builds open-weight agentic coding models you can self-host for secure, long-horizon software engineering.

MiniMax M3 brings 1M-token context, native multimodality, and frontier coding to a monthly token subscription.

Open-source multimodal model with native tool use for building autonomous agents that see and act.

ModelScope is Alibaba Cloud's open-source Model-as-a-Service hub for finding, fine-tuning, and deploying AI models.

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

LFM2.5 is Liquid AI's open-weight on-device AI family, running native text, vision, and audio models locally on CPU, GPU, or NPU.

Baichuan-7B is a free open-source 7B bilingual Chinese-English base LLM on Hugging Face for fine-tuning and research.

Mistral's open-weight coding model family — Devstral 2 (123B) and Devstral Small 2 (24B) — paired with the open-source Mistral Vibe CLI for terminal-native

Kimi K2.6 is an open-source coding model built for long-horizon agentic coding and 300-agent swarm orchestration.

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