Cloudkart.ai
Z.ai logo

Z.ai

Open Source

A free chat assistant from Z.ai (formerly Zhipu), backed by its own GLM models released under a permissive MIT license. GLM-4.6 handles a 200K-token context and performs well on coding and agentic tasks, and is a common drop-in for tools like Claude Code, Cline and Roo Code.

ai assistantchatbotopen sourceglmcodingagent mode
Visit Z.ai

Work at Z.ai? Manage this listing

Our take

Z.ai (formerly Zhipu) is a free chat assistant backed by its own GLM models under a permissive MIT license. GLM-4.6 handles a 200K-token context and does well on coding and agentic tasks - a common drop-in for Claude Code, Cline and Roo Code. Free to use and self-hostable; strong value without a subscription. China-hosted, which some teams must weigh.

Best for

Developers and power users who want a capable, low-cost frontier assistant - especially for coding and agentic workflows - or open weights they can self-host.

Pros

  • Capable GLM-4.6 model with a 200K-token context, strong at coding and agentic tasks
  • Free chat plus MIT-licensed open weights you can self-host
  • Works as a cheaper drop-in model for Claude Code, Cline, Roo Code and Kilo Code
  • Backed by a real, now publicly listed company (HK IPO, Jan 2026)

Cons

  • China-hosted service - a data-governance question for some organisations
  • Brand and ecosystem are less familiar to global users than OpenAI or Anthropic
  • Newer entrant; behaviour and limits still settling

How it compares

Like DeepSeek, Kimi and Qwen, Z.ai offers a strong, low-cost alternative to Western assistants; its edge is the permissive MIT licence on flagship GLM models and solid coding/agent performance.

Full review

Z.ai is the consumer-facing assistant from the company formerly called Zhipu AI, a 2019 spinout from Tsinghua University that listed on the Hong Kong exchange in January 2026. It runs on the company's own GLM family - the current GLM-4.6 carries a 200K-token context and posts competitive coding and agentic-task scores.

What makes it interesting beyond the chat box is licensing: GLM weights have shipped under the MIT licence since mid-2025, so teams can self-host or wire the model into agent stacks like Claude Code, Cline and Roo Code at a fraction of frontier pricing. The main thing to weigh is governance - it is a China-hosted service, which some organisations will need to assess before sending sensitive data.

Cloudkart Trust Graph

3.8/5
  • Actual Utility4/5

    Source: Initial LLM-authored rubric (backfill)

  • Ease of Use4/5

    Source: Initial LLM-authored rubric (backfill)

  • Pricing Fairness5/5

    Source: Initial LLM-authored rubric (backfill)

  • Reliability3/5

    Source: Initial LLM-authored rubric (backfill)

  • Differentiation3/5

    Source: Initial LLM-authored rubric (backfill)

Scored as of . Each score is versioned and auditable; vendors cannot buy it.

How this score is set

Editorial rubric
Primary signal — five dimensions, 3.8/5 average.
Community reviews
None yet.
Pricing verified
Not yet verified
Independence
Score set by our editorial team before any affiliate relationship is considered. No vendor can buy it.

How we keep this independent →

Frequently asked questions

Is Z.ai free, and how much does it cost?
Z.ai is open source and free to self-host.
Who is Z.ai best for?
Developers and power users who want a capable, low-cost frontier assistant - especially for coding and agentic workflows - or open weights they can self-host.
How is Z.ai rated on Cloudkart.ai?
Z.ai scores 3.8 out of 5 on the Cloudkart.ai rubric, which weighs actual utility, ease of use, pricing fairness, reliability and differentiation. Scores are set editorially and can never be bought.

Community reviews

No community reviews yet. Be the first to share how Z.ai works for you.

Relevant tools

More tools in Writing & Content Generation.

Compare Z.ai head-to-head: vs Perplexity · vs Claude · vs Google Gemini · vs Grok