Most reviews evaluating AI fashions are based mostly on benchmarks of efficiency, however a current analysis report from Sonar takes a special strategy: grouping completely different fashions by their coding personalities and searching on the downsides of every in the case of code high quality.
The researchers studied 5 completely different LLMs utilizing the SonarQube Enterprise static evaluation engine on over 4,000 Java assignments. The LLMs reviewed had been Claude Sonnet 4, OpenCoder-8B, Llama 3.2 90B, GPT-4o, and Claude Sonnet 3.7.
They discovered that the fashions had completely different traits, comparable to Claude Sonnet 4 being very verbose in its outputs, producing over 3x as many strains of code as OpenCoder-8B for a similar drawback.
Primarily based on these traits, the researchers divided the 5 fashions into coding archetypes. Claude Sonnet 4 was the “senior architect,” writing refined, complicated code, however introducing high-severity bugs. “Due to the extent of technical problem tried, there have been extra of those points,” stated Donald Fischer, a VP at Sonar.
OpenCoder-8B was the “speedy prototyper” on account of it being the quickest and most concise whereas additionally doubtlessly creating technical debt, making it supreme for proof-of-concepts. It created the best subject density of all of the fashions, with 32.45 points per thousand strains of code.
Llama 3.2 90B was the “unfulfilled promise,” as its scale and backing implies it needs to be a top-tier mannequin, but it surely solely had a move charge of 61.47%. Moreover, 70.73% of the vulnerabilities it created had been “BLOCKER” severity, essentially the most extreme kind of bug, which prevents testing from persevering with.
GPT-4o was an “environment friendly generalist,” a jack-of-all-trades that may be a widespread selection for general-purpose coding help. Its code wasn’t as verbose because the senior architect or as concise because the speedy prototyper, however someplace within the center. It additionally averted producing extreme bugs for essentially the most half, however 48.15% of its bugs had been control-flow errors.
“This paints an image of a coder who accurately grasps the primary goal however usually fumbles
the main points required to make the code sturdy. The code is prone to perform for the meant situation however shall be affected by persistent issues that compromise high quality and reliability over time,” the report states.
Lastly, Claude 3.7 Sonnet was a “balanced predecessor.” The researchers discovered that it was a succesful developer that produced well-documented code, however nonetheless launched a lot of extreme vulnerabilities.
Although the fashions did have these distinct personalities, additionally they shared related strengths and weaknesses. The widespread strengths had been that they rapidly produced syntactically appropriate code, had stable algorithmic and knowledge construction fundamentals, and effectively translated code to completely different languages. The widespread weaknesses had been that all of them produced a excessive proportion of high-severity vulnerabilities, launched extreme bugs like useful resource leaks or API contract violations, and had an inherent bias in direction of messy code.
“Like people, they turn into inclined to refined points within the code they generate, and so there’s this correlation between functionality and threat introduction, which I believe is amazingly human,” stated Fischer.
One other attention-grabbing discovering of the report is that newer fashions could also be extra technically succesful, however are additionally extra prone to generate dangerous code. For instance, Claude Sonnet 4 has a 6.3% enchancment over Claude 3.7 Sonnet on benchmark move charges, however the points it generated had been 93% extra prone to be “BLOCKER” severity.
“If you happen to suppose the newer mannequin is superior, give it some thought another time as a result of newer is just not truly superior; it’s injecting an increasing number of points,” stated Prasenjit Sarkar, options advertising supervisor at Sonar.
How reasoning modes impression GPT-5
The researchers adopted up their report this week with new knowledge on GPT-5 and the way the 4 accessible reasoning modes—minimal, low, medium, and excessive—impression efficiency, safety, and code high quality.
They discovered that rising reasoning has a diminishing return on useful efficiency. Bumping up from minimal to low ends in the mannequin’s move charge rising from 75% to 80%, however medium and excessive solely had a move charge of 81.96% and 81.68%, respectively.
By way of safety, excessive and low reasoning modes remove widespread assaults like path-traversal and injection, however substitute them with harder-to-detect flaws, like insufficient I/O error-handling. The low reasoning mode had the best proportion of that subject at 51%, adopted by excessive (44%), medium (36%), and minimal (30%).
“We have now seen the path-traversal and injection turn into zero p.c,” stated Sarkar. “We will see that they’re attempting to resolve one sector, and what’s taking place is that whereas they’re attempting to resolve code high quality, they’re someplace doing this trade-off. Insufficient I/O error-handling is one other drawback that has skyrocketed. If you happen to have a look at 4o, it has gone to 15-20% extra within the newer mannequin.”
There was an identical sample with bugs, with control-flow errors lowering past minimal reasoning, however superior bugs like concurrency / threading rising alongside the reasoning problem.
“The trade-offs are the important thing factor right here,” stated Fischer. “It’s not as simple as to say, which is one of the best mannequin? The way in which this has been seen within the horse race between completely different fashions is which of them full essentially the most variety of options on the SWE-bench benchmark. As we’ve demonstrated, the fashions that may do extra, that push the boundaries, additionally they introduce extra safety vulnerabilities, they introduce extra maintainability points.”