CodePanel

DeepSeek V4 Pro (Workers AI) vs Gemini 3.8 Flash

Gemini 3.8 Flash reads a 26.2K-token pull request for $0.0197, 43% less than DeepSeek V4 Pro (Workers AI) ($0.0346). From 2027-01-01, after a scheduled reprice, DeepSeek V4 Pro (Workers AI) becomes the cheaper one. 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: DeepSeek V4 Pro (Workers AI) minus Gemini 3.8 Flash is −14 pts (95% interval −29 pts to 0 pts).

Side by side

DeepSeek V4 Pro (Workers AI) and Gemini 3.8 Flash side by side
FigureDeepSeek V4 Pro (Workers AI)Gemini 3.8 Flash
Price tierMid-rangeMid-range
Bugs caughtshare of known bugs54%68% (better)
False alarmsper clean PR2.91.0 (better)
Cost per reviewmeasured$0.0643 (better)$0.0862
Cost per bug caught$0.0872 (better)$0.1166
Exact linecatches on the expected lines89%94% (better)
Review timemean390s (better)427s
Typical review calllist price$0.0442$0.0287 (better)
Input / 1M tokens$1.32$0.75 (better)
Output / 1M tokens$3.96$3.75 (better)
Context window1M1M
  • From 2027-01-01, after a scheduled reprice, DeepSeek V4 Pro (Workers AI) is the cheaper of the two.

Head to head

The known bugs both were scored on, by who caught each in at least one run.

Of 42 known bugs both were scored on: Both 25Only DeepSeek V4 Pro (Workers AI) 9Only Gemini 3.8 Flash 6Neither 2

By language

Share of known bugs each model caught, by language
ModelAllTypeScript6 bugsGo4 bugsJava4 bugsRust4 bugsC#3 bugsJavaScript3 bugsKotlin3 bugsPHP3 bugsPython3 bugsRuby3 bugsSwift3 bugsC++2 bugsC1 bug
DeepSeek V4 Pro (Workers AI)54%79%25%49%54%59%17%30%69%72%67%75%38%33%
Gemini 3.8 Flash68%83%75%43%75%78%67%0%50%83%100%100%50%50%

Where they differ

Of the 42 known bugs both were scored on, both caught 25 and neither caught 2. DeepSeek V4 Pro (Workers AI) caught 9 bugs that Gemini 3.8 Flash missed: 3 in Kotlin, 2 in Java, 1 in C++, 1 in JavaScript, 1 in PHP and 1 in Rust. Gemini 3.8 Flash caught 6 bugs that DeepSeek V4 Pro (Workers AI) missed: 2 in Go, 2 in JavaScript, 1 in C# and 1 in Rust.

On pull requests with no known bug, DeepSeek V4 Pro (Workers AI) raised 2.9 false alarms per review and Gemini 3.8 Flash 1. A measured review cost $0.0643 on DeepSeek V4 Pro (Workers AI), 25% less than $0.0862 on Gemini 3.8 Flash.

Each known bug, and how often each model caught it
Pull requestLanguage · kindDeepSeek V4 Pro (Workers AI)Gemini 3.8 Flash
nushell/nushellreorganize nushell configurationRust · Error handling3/40/1
jhy/jsoupStore cached child els in Attributes, not fieldJava · Concurrency3/60/4
kotest/kotestFix DataClassEq comparatorKotlin · Logic2/40/1
nodejs/undicifeat: add configurable maxPayloadSize for WebSocketJavaScript · Security2/40/1
symfony/symfony[Security] Allow multiple OIDC discovery endpointsPHP · Security1/30/1
ggml-org/llama.cpprpc : cache and reuse compute graphsC++ · Security1/40/1
kotest/kotestRefactor eventually to ensure wall-clock time usageKotlin · Error handling1/40/1
mockito/mockitoAdd core API to enable Kotlin singleton mockingJava · API misuse1/50/1
ktorio/ktorKTOR-9171 Redesign ByteReadChannel.readUTF8Line APIKotlin · Logic1/70/1
denoland/denoperf(libs/core): poll only ready handles and yield between I/O batchesRust · Concurrency0/41/1
dotnet/orleansfix(serialization): cap codec allocationsC# · Logic0/41/1
go-chi/chifeat: middleware.ClientIP, a replacement for middleware.RealIPGo · Security0/41/1
golang/gocrypto/tls: add LocalCertificate to ConnectionStateGo · Logic0/42/2
pinojs/thread-streamrefactor: replace setTimeout with Atomics.waitAsyncJavaScript · Concurrency0/42/2
sindresorhus/p-throttleAdd `weight` optionJavaScript · Logic0/41/1

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: DeepSeek V4 Pro (Workers AI) and Gemini 3.8 Flash.