Gemini 3.8 Flash vs Gemini 4 Argon
Reading a 26.2K-token pull request costs $0.0524 on Gemini 4 Argon, 2.7× the $0.0197 it costs on Gemini 3.8 Flash. Both sit in the mid-range price tier, so the choice between them is about review quality more than budget. The benchmark cannot separate their recall: Gemini 3.8 Flash minus Gemini 4 Argon is −14 pts (95% interval −27 pts to −2 pts).
Side by side
| Figure | Gemini 3.8 Flash | Gemini 4 Argon |
|---|---|---|
| Price tier | Mid-range | Mid-range |
| Bugs caughtshare of known bugs | 68% | 82% (better) |
| False alarmsper clean PR | 1.0 (better) | 1.1 |
| Cost per reviewmeasured | $0.0862 | $0.0527 (better) |
| Cost per bug caught | $0.1166 | $0.0706 (better) |
| Exact linecatches on the expected lines | 94% (better) | 78% |
| Review timemean | 427s | 99s (better) |
| Typical review calllist price | $0.0287 (better) | $0.0766 |
| Input / 1M tokens | $0.75 (better) | $2.00 |
| Output / 1M tokens | $3.75 (better) | $10.00 |
| Context window | 1M | 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 3.8 Flash | 68% | 83% | 75% | 43% | 75% | 78% | 67% | 0% | 50% | 83% | 100% | 100% | 50% | 50% |
| Gemini 4 Argon | 82% | 83% | 50% | 83% | 75% | 89% | 67% | 100% | 100% | 83% | 100% | 100% | 100% | 0% |
Of the 42 known bugs both were scored on, both caught 29 and neither caught 4. Gemini 3.8 Flash caught 2 bugs that Gemini 4 Argon missed: 1 in C and 1 in Go. Gemini 4 Argon caught 7 bugs that Gemini 3.8 Flash missed: 3 in Kotlin, 2 in Java, 1 in C++ and 1 in PHP.
On pull requests with no known bug, Gemini 3.8 Flash raised 1 false alarms per review and Gemini 4 Argon 1.1. A measured review cost $0.0862 on Gemini 3.8 Flash, 64% more than $0.0527 on Gemini 4 Argon.
| Pull request | Language · kind | Gemini 3.8 Flash | Gemini 4 Argon |
|---|---|---|---|
| golang/gocrypto/tls: add LocalCertificate to ConnectionState | Go · Logic | 2/2 | 0/1 |
| c-ares/c-aresEvent Thread could stall when not notified of new queries on existing connections that are in a bad state | C · Resource handling | 1/2 | 0/1 |
| ggml-org/llama.cpprpc : cache and reuse compute graphs | C++ · Security | 0/1 | 1/1 |
| jhy/jsoupStore cached child els in Attributes, not field | Java · Concurrency | 0/4 | 1/2 |
| kotest/kotestFix DataClassEq comparator | Kotlin · Logic | 0/1 | 1/1 |
| kotest/kotestRefactor eventually to ensure wall-clock time usage | Kotlin · Error handling | 0/1 | 1/1 |
| ktorio/ktorKTOR-9171 Redesign ByteReadChannel.readUTF8Line API | Kotlin · Logic | 0/1 | 1/1 |
| mockito/mockitoAdd core API to enable Kotlin singleton mocking | Java · API misuse | 0/1 | 1/1 |
| symfony/symfony[Security] Allow multiple OIDC discovery endpoints | PHP · Security | 0/1 | 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: Gemini 3.8 Flash and Gemini 4 Argon.