Editorial illustration for GLM-5.3 API Launch Priced at USD 1.4/USD 4.4 Per Million Tokens
GLM-5.3 API Launches at $1.4-$4.4 Per Million Tokens
Z.ai's GLM-5.3 landed on the API this week, a week after the Chinese startup's model reportedly turned up a previously undetected vulnerability in Cursor during testing. Developers can now build directly on top of it, wiring the model into agents and applications rather than waiting on the coming weights release. Z.ai says it still plans to open-source those weights, though it hasn't set a date or picked a license.
For now, anyone on a GLM Coding Plan is stuck with the OpenAI Chat Completions-compatible protocol, which limits some integration options. But the number drawing attention is the price: $1.40 per million input tokens and $4.40 per million output tokens, identical to what GLM-5.2 charged. Cached input runs $0.26 per million tokens, and Z.ai is waiving cached-input storage fees for now.
That pricing decision matters because Z.ai is framing GLM-5.3 as a real jump in coding and long-horizon agent performance, not just an incremental update. Holding the rate steady means developers get whatever gains the new model offers without eating a price hike, at a moment when rival labs are settling into some very different rate structures.
For developers, though, the immediate change is straightforward: GLM-5.3 is now callable through Z.ai’s API at the same $1.40/$4.40 per-million-token rate as GLM-5.2, giving teams another relatively low-cost option for testing frontier-class coding and agent workloads.
Why this matters
The math here is blunt: $1.4 in, $4.4 out per million tokens undercuts Gemini 3.7 Flash's promotional $0.75/$3.75/$4.50 tiered pricing only in some spots and loses in others, so z.ai isn't winning on price alone. It's winning on the story around price: an open-source frontier model that already found a real vulnerability in Cursor before it even had a public API rate card. For developers weighing which model to route traffic through, that's the actual signal, not the per-token math.
Founders building on GLM-5.3 get an open license and a security narrative that's hard to buy at any price. Researchers should treat the Cursor vulnerability claim as the thing worth verifying independently, since a single unconfirmed find is doing a lot of marketing work right now. Google's move to lock Gemini Flash pricing through the end of 2026 tells you the incumbents feel the pressure.
Watch whether z.ai's pricing holds once promotional windows close elsewhere, and whether that cyber capability shows up again outside a press cycle.
Related Reading
Open Source D&B rebuilds 642 million‑business database after AI agents hit limits
Open Source OpenAI's GPT Transcribe Cuts Error Rate to 3.31%, Improving on GPT-4o
Open Source Study: 6.7% of Deepfake Requests on Hugging Face Targeted Children
From Archives Block's Berd Agent Works Across AI Models, Stores History Locally
From Archives NVIDIA's TensorRT Model Connect Streamlines Hugging Face to C++ Inference
Common Questions Answered
What is the pricing structure for GLM-5.3 API compared to GLM-5.2?
GLM-5.3 is available through Z.ai's API at the same $1.40/$4.40 per-million-token rate as GLM-5.2, maintaining consistent pricing for developers. This pricing provides a relatively low-cost option for testing frontier-class coding and agent workloads compared to competing models.
What vulnerability did GLM-5.3 discover during testing before its public API launch?
During testing, GLM-5.3 reportedly discovered a previously undetected vulnerability in Cursor that had not been found before. This discovery demonstrates the model's capability for identifying security issues and adds credibility to its frontier-class performance.
Will Z.ai release the GLM-5.3 model weights as open-source?
Yes, Z.ai has stated it still plans to open-source the GLM-5.3 weights, though the company has not yet set a specific release date or chosen a license for the open-source release. Developers can currently access the model through the API while waiting for the weights to become available.
How does GLM-5.3's pricing compare to Gemini 3.7 Flash's promotional pricing?
GLM-5.3's $1.40/$4.40 per-million-token pricing undercuts Gemini 3.7 Flash's promotional $0.75/$3.75/$4.50 tiered pricing in some scenarios but loses in others. However, Z.ai's competitive advantage lies not primarily in price but in the model's demonstrated ability to find real vulnerabilities like the Cursor discovery.
Further Reading
- GLM-5.3 hits the API at $1.4/$4.4 per million tokens - VentureBeat
- Z.ai launches GLM-5.3 API for coding and defensive security - RuntimeWire
- GLM-5.3 identifies serious vulnerability in Cursor code editor - CryptoBriefing
- What Is GLM-5.3? Zhipu's Open-Weight Coding Model ... - Apidog
- GLM-5.3: Z.ai's New Open Model - El Solitario
