Crypto news

A buzz has erupted across crypto and tech communities around the new open-source neural network GLM-5.2 from Z.ai. Enthusiasts and bloggers are already calling it a "killer" of Anthropic's flagship model, Claude Opus 4.8, and some claim it outperforms the competitor in several scenarios at a ten times lower price. Let's figure out how justified these bold claims are.

What is GLM-5.2 and what is its main advantage?

The developers position GLM-5.2 as a flagship model tailored for long working sessions. The key difference from its predecessor GLM-5.1 is a stable context window of 1 million tokens (compared to 200 thousand). This allows the model to keep vast amounts of code and text in its "field of view" without losing quality on ultra-long tasks.

Key features:

• 1 million token context without degradation on ultra-long sessions.
• Two levels of reasoning effort: High (balance of performance and token consumption) and Max (maximum capability).
• Open-source MIT license with no regional restrictions, allowing the model to be run on your own hardware (self-hosting).
• API pricing remains at the level of the previous GLM-5.1 version.

The model is already available on HuggingFace and ModelScope, as well as through the GLM Coding Plan subscription and the ZCode desktop agent.

Benchmarks: Where GLM-5.2 excels and where it falls short

According to Z.ai's own tests, GLM-5.2 is recognized as the strongest open-source model on the market. However, it generally falls short of Anthropic Claude Opus 4.8 in most cases.

Here are the results on key tests in Max mode (maximum reasoning):

On long-horizon tasks, the picture is similar. On the FrontierSWE test, where the model manages open technical projects for tens of hours, GLM-5.2 lags behind Opus 4.8 by only 1%, but outperforms GPT-5.5 and the previous version Opus 4.7. On PostTrainBench, it also surpasses Opus 4.7 and GPT-5.5, yielding only to Opus 4.8.

Price and the Catch: Cheap, but Not Without Nuances

The GLM Coding Plan subscription is divided into three tiers with annual discounts of 30%:

• Lite: $12.6/month (instead of $18).
• Pro: $50.4/month (instead of $72).
• Max: $112/month (instead of $160).

Within the subscription, quota consumption depends on load: a 3x multiplier during peak hours (14:00-18:00 Beijing time) and 2x off-peak. Until the end of September, a promotion applies—off-peak usage is billed at 1x.

However, users note that the model only reveals its full potential in Max mode, which consumes significantly more tokens than High mode. This makes its use more expensive than it initially appears.

User Reviews: Enthusiasm vs. Criticism

Strengths:

• The model is called the strongest open-source neural network tested so far.
• Basic logic is noticeably better than version 5.1, and in programming, the model is comparable to GPT-5.5 at a high reasoning level.
• The AI autonomously performs complex tasks through auxiliary agents and proactively suggests fixing inconsistencies it notices.
• Users describe it as slow and expensive, but extremely persistent in achieving its goal.

Criticism:

• The cloud infrastructure, despite a good mathematical model, is described as extremely weak.
• Developers complain about expensive pricing and poor support, noting it's easier to pay for Claude or GPT.
• The neural network is criticized for its tendency to get stuck in infinite loops and ignore commands. According to users, the model is tailored exclusively for benchmarks.

Conclusion: Killer or Competitor?

There is no clear answer. GLM-5.2 is the best open-source model available today for programming and autonomous tasks. In certain long scenarios, it comes very close to Anthropic's flagship. The open-source MIT license, the ability to run on your own hardware, and the low entry barrier make it a notable player.

However, it is bloggers, not benchmarks, who call it a "killer" of Claude. According to most of Z.ai's own tests, their model ranks below Opus 4.8. Furthermore, users complain about unstable cloud infrastructure, high token consumption in Max mode, and poor support.

My expert opinion: GLM-5.2 is a significant step forward for open-source models, but it is still far from fully "killing" Claude. It narrows the gap with the leaders but does not yet surpass them. If you are willing to tolerate the current infrastructure shortcomings and high token consumption, it is an excellent tool for autonomous programming. For mass use, it is still too early.