Z.ai Pricing in 2026: The Free Chat App, GLM API Costs and the GLM Coding Plan
What Z.ai costs in 2026: the free Z.ai chat app, per-token prices for every GLM model on the API, the free GLM Flash models, and the GLM Coding Plan tiers, compared with other providers that serve GLM.
The Z.ai chat app is free: you can use it without an account, with GLM-5.3-Flash as the default model and GLM-5.3 one click away. On the API, the flagship GLM-5.3 costs $1.40 per million input tokens and $4.40 per million output tokens, and three smaller models (including GLM-4.7-Flash) are priced at $0. If you want GLM inside Claude Code, Cursor or another coding agent, the GLM Coding Plan starts at $18 a month, or $12.60 a month billed yearly.
Z.ai is the international brand of Zhipu AI, the Chinese lab behind the GLM models. Prices below come from Z.ai’s official pricing and subscription pages and our own provider data, as of November 2026.
Is Z.ai free?
Yes, in two ways.
- The chat app at chat.z.ai is free. It opens in guest mode with no sign-in, and the model picker offers GLM-5.3-Flash, GLM-5.3 and GLM-5.2. We didn’t find a paid chat subscription on Z.ai’s site, and it doesn’t publish message limits for free use.
- The API has free models. GLM-4.7-Flash, GLM-4.5-Flash and GLM-4.6V-Flash (vision) are listed at $0 for input, cached input and output. They’re the “Z.ai free tier” most people are looking for: real API access at no cost, on smaller models.
The current flagship and Flash models are paid on the API. Z.ai’s pricing page doesn’t mention sign-up credits.
How much does the GLM API cost?
Z.ai bills per million tokens, pay-as-you-go. Cached input is billed at about a fifth of the normal input rate, and cache storage is listed as free for a limited time.
| Model | Input / 1M | Cached input / 1M | Output / 1M |
|---|---|---|---|
| GLM-5.3 | $1.40 | $0.26 | $4.40 |
| GLM-5.3-FlashX | $0.37 | $0.075 | $1.25 |
| GLM-5.3-Flash | $0.15 | $0.03 | $0.50 |
| GLM-5.2 | $1.40 | $0.26 | $4.40 |
| GLM-5.1 | $1.40 | $0.26 | $4.40 |
| GLM-5 | $1.00 | $0.20 | $3.20 |
| GLM-4.7 | $0.60 | $0.11 | $2.20 |
| GLM-4.6V (vision) | $0.30 | $0.05 | $0.90 |
| GLM-4.5-Air | $0.20 | $0.03 | $1.10 |
| GLM-4.7-Flash | Free | Free | Free |
A few things stand out. GLM-5.1, 5.2 and 5.3 all cost the same, so there’s no saving in staying on an older version. GLM-5.3-Flash is the value pick at $0.15 / $0.50 with a 1M-token context window, roughly a ninth of the flagship’s price.
Beyond text, Z.ai charges $0.01 per image for CogView-4, $0.015 per image for GLM-Image, $0.20 per video for CogVideoX-3, and $0.01 per call for its built-in web search tool.
What is the GLM Coding Plan and how much is it?
The GLM Coding Plan is a subscription for using GLM inside coding and agent tools rather than paying per token. Z.ai lists support for 20+ tools, including Claude Code, Codex, Cursor, OpenCode, Cline and its own ZCode IDE.
| Plan | Monthly | Quarterly (20% off) | Yearly (30% off) | Credits per 5 hours | Credits per week |
|---|---|---|---|---|---|
| Lite | $18 | $14.40/mo | $12.60/mo | 2,000 | 10,000 |
| Pro | $80 | $64/mo | $56/mo | 12,000 | 60,000 |
| Max | $168 | $134.40/mo | $117.60/mo | 28,000 | 140,000 |
How it works:
- Models: GLM-5.3 and GLM-5.3-Flash. Requests for GLM-5.2 or 5.1 are routed to GLM-5.3, and GLM-4.7 requests go to GLM-5.3-Flash.
- Two limits at once. A rolling 5-hour credit quota and a weekly quota. When you hit one, you wait for the reset; your account balance isn’t charged.
- Off-peak discount. Usage outside peak hours (Monday to Friday, 14:00 to 18:00 Singapore time) costs half the usual credits.
- Tools only. The plan works through supported tools. Z.ai’s FAQ says direct API calls outside the plan aren’t available, so for your own app you still pay API rates.
- No refunds. Subscriptions are non-refundable; cancel at least 24 hours before renewal to stop auto-renewal.
Pro adds faster generation and a set of MCP tools; Max adds dedicated resources at peak times. Team plans are also offered.
Is it worth it? At API rates, $18 buys about 4 million GLM-5.3 output tokens, or about 13 million input tokens. Coding agents re-send large chunks of your codebase on every turn, so token counts climb fast with daily use. If you code with an agent most days, a fixed monthly plan is usually cheaper and easier to budget. For occasional use, or for anything outside a coding tool, pay-as-you-go API access is simpler.
Is GLM cheaper outside Z.ai?
GLM models are open weight, so many providers host them. Most match Z.ai’s price exactly, but a few go lower, usually with quantized weights or slower tiers.
| Model | Z.ai | Cheaper options we track |
|---|---|---|
| GLM-5.3 | $1.40 / $4.40 | DeepInfra $0.90 / $4.00; OpenRouter lists a 4-bit endpoint at $0.07 / $1.40 |
| GLM-5.2 | $1.40 / $4.40 | OpenRouter from $0.28 / $0.88; DeepInfra $0.75 / $2.40 |
| GLM-5 | $1.00 / $3.20 | OpenRouter $0.60 / $1.92; DeepInfra $0.60 / $2.08 |
| GLM-4.7 | $0.60 / $2.20 | OpenRouter from $0.32 / $1.40 |
| GLM-5.3-Flash | $0.15 / $0.50 | Same price at Together AI, Fireworks and most others |
OpenRouter’s cheapest endpoint isn’t always the one its default routing picks, so pin the provider if price matters. Hosted copies of GLM-4.7-Flash cost around $0.06 / $0.40 elsewhere, which makes Z.ai’s free tier the obvious choice for that model. Cloudflare Workers AI also has a free allowance (10,000 Neurons a day) that covers light use of GLM-4.7-Flash; its newer GLM models need a paid billing method.
For a wider view, see the cheapest LLM APIs and free LLM APIs rankings.
Which Z.ai option should you pick?
- Just want to chat: use the free Z.ai chat app. Other Chinese chat apps we track include Qwen Chat, DeepSeek and Kimi.
- Building on a budget: start with free GLM-4.7-Flash, then move to GLM-5.3-Flash at $0.15 / $0.50.
- Need the flagship in production: GLM-5.3 at $1.40 / $4.40 from Z.ai, or a cheaper host such as DeepInfra if you can accept a third party.
- Coding every day in Claude Code or Cursor: the GLM Coding Plan, Lite or Pro, billed quarterly or yearly.
For background on the model family, see our posts on GLM-5.2’s coding results and China’s open-weight frontier models. Prices change often, so check the Zhipu AI provider page for current figures.