GLM-5.3 (Workers AI) vs GPT-6.1 Sol
GLM-5.3 (Workers AI) reads a 26.2K-token pull request for $0.0367, 30% less than GPT-6.1 Sol ($0.0524). Both sit in the mid-range price tier, so the choice between them is about review quality more than budget. In the benchmark GPT-6.1 Sol caught 28 percentage points more of the known defects than GLM-5.3 (Workers AI) (95% interval −42 pts to −14 pts, a − b).
Side by side
| Figure | GLM-5.3 (Workers AI) | GPT-6.1 Sol |
|---|---|---|
| Price tier | Mid-range | Mid-range |
| Bugs caughtshare of known bugs | 61% | 89% (better) |
| False alarmsper clean PR | 1.9 | 0.9 (better) |
| Cost per reviewmeasured | $0.0311 (better) | $0.0401 |
| Cost per bug caught | $0.0559 | $0.0514 (better) |
| Exact linecatches on the expected lines | 85% | 95% (better) |
| Review timemean | 65s | 28s (better) |
| Typical review calllist price | $0.0473 (better) | $0.0766 |
| Input / 1M tokens | $1.40 (better) | $2.00 |
| Output / 1M tokens | $4.40 (better) | $10.00 |
| Context window | 1M | 1.1M |
Head to head
The known bugs both were scored on, by who caught each in at least one run.
By language
| Model | All | TypeScript6 bugs | Go4 bugs | Java4 bugs | Rust4 bugs | C#3 bugs | JavaScript3 bugs | Kotlin3 bugs | PHP3 bugs | Python3 bugs | Ruby3 bugs | Swift3 bugs | C++2 bugs | C1 bug |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GLM-5.3 (Workers AI) | 61% | 78% | 19% | 46% | 73% | 48% | 33% | 72% | 44% | 58% | 67% | 100% | 100% | 67% |
| GPT-6.1 Sol | 89% | 83% | 75% | 70% | 100% | 89% | 100% | 100% | 100% | 100% | 100% | 100% | 50% | 100% |
Of the 42 known bugs both were scored on, both caught 33 and neither caught 1. GLM-5.3 (Workers AI) caught 3 bugs that GPT-6.1 Sol missed: 1 in C++, 1 in Java and 1 in TypeScript, most of them concurrency bugs. GPT-6.1 Sol caught 5 bugs that GLM-5.3 (Workers AI) missed: 1 in C#, 1 in Go, 1 in JavaScript, 1 in PHP and 1 in Ruby, most of them logic bugs.
On pull requests with no known bug, GLM-5.3 (Workers AI) raised 1.9 false alarms per review and GPT-6.1 Sol 0.9. A measured review cost $0.0311 on GLM-5.3 (Workers AI), 23% less than $0.0401 on GPT-6.1 Sol.
| Pull request | Language · kind | GLM-5.3 (Workers AI) | GPT-6.1 Sol |
|---|---|---|---|
| ggml-org/llama.cpprpc : cache and reuse compute graphs | C++ · Security | 3/3 | 0/1 |
| trpc/trpcfix(server): pass maxDurationMs timeout signal to subscription handlers | TypeScript · Concurrency | 2/3 | 0/1 |
| jhy/jsoupStore cached child els in Attributes, not field | Java · Concurrency | 2/6 | 0/2 |
| dotnet/orleansfix(serialization): cap codec allocations | C# · Logic | 0/3 | 1/1 |
| go-chi/chifeat: middleware.ClientIP, a replacement for middleware.RealIP | Go · Security | 0/3 | 1/1 |
| rails/railsReimplement Action Cable redis adapter with redis-client | Ruby · Logic | 0/3 | 1/1 |
| sindresorhus/p-throttleAdd `weight` option | JavaScript · Logic | 0/3 | 1/1 |
| symfony/symfony[Security] Allow multiple OIDC discovery endpoints | PHP · Security | 0/3 | 1/1 |
Defects caught over all runs: a PR reviewed three times with one known bug shows 3/3 when it was caught every time.
Every bug each was scored on: GLM-5.3 (Workers AI) and GPT-6.1 Sol.