cd /news/ai-research/gemini-3-1-pro-on-iac-bench-2026-10 · home › topics › ai-research › article
[ARTICLE · art-173575] src=massdriver.cloud ↗ pub= topic=ai-research verified=true sentiment=· neutral

Gemini 3.1 Pro on iac-bench 2026-10

Gemini 3.1 Pro scored an overall composite of 73.0 (95% CI 69.2–76.2) on Massdriver's iac-bench 2026-10 infrastructure-as-code benchmark, ranking 13th of 16 models, with 5 of its 9 OpenTofu attempts flagged for security findings. All 9 modules planned and none failed a gate, but the model never set `https_only`, basic-auth publishing flags or `ssl_mode`, leaving HTTP and shared-credential publishing open, and it was the only model with no `validation` block in any of its nine modules. The benchmark, run on 2026-10-22 generation and 2026-10-23 judging at a list price of $2.00/$12.00 per million input/output tokens, cost $2.26 for all 9 attempts.

recommendation

Score insecure IaC configuration with rubric rules, not an LLM judge: on Cloud SQL the attempt with a security finding got the highest judge grade, so combine both components before ranking models.

more in playbook: Evals and testing for LLM apps
by read3 min views6 publishedNov 4, 2026
Gemini 3.1 Pro on iac-bench 2026-10
Image: Massdriver (auto-discovered)

New: LLM Rankings are live!

See how the top LLMs rank on real infrastructure tasks

How Gemini 3.1 Pro did on iac-bench 2026-10: an overall composite of 73.0, scores per fixture, strengths, weaknesses and the patterns in its OpenTofu modules.

Every model evaluated on the 2026-10 fixtures side by side, with a sortable leaderboard, picks by use case and the cost of each: LLM rankings.

| Developer | | | API identifier | gemini-3.1-pro-preview | | Access | Gemini API | | Weights | Closed | | List price (USD / M tokens, in / out) | 2.00 / 12.00 (OpenRouter, 2026-10-23) | | Evaluated | 2026-10-22 (generation) · 2026-10-23 (judge) | | Attempts | 9 (3 fixtures × 3 samples) | | Overall composite | 73.0 (95% CI 69.2–76.2) | | Rank | 13 of 16 (95% rank interval 11–15) | | Mechanical only / judge only | 71.9 (rank 13) / 71.9 (rank 14) | | Failed gates · insecure attempts | 0/9 · 5/9 | | Estimated cost, all 9 attempts | $2.26 |

Gemini 3.1 Pro writes the leanest modules among the full-size models: few inputs, short files, low token use. All 9 of its modules planned. Five of the 9 had security findings, all from settings left unset rather than settings exposed as inputs:

https_only and the basic-auth publishing flags are never set, so the provider defaults leave HTTP and shared-credential publishing open.ssl_mode is never set, so unencrypted connections are accepted. It is also the only model with no validation block in any of its nine modules.

Fixture Composite median (range) Rubric median Expertise Judge median ✗ gate ⚠ security Median inputs Median output tokens Median time Median cost
G001 · S3 74.2 (73.4–80.1) 0.733 3/7 78.9 0/3 0/3 5 4,677 140 s $0.171
G002 · Cloud SQL 79.3 (73.5–82.0) 0.815 3/9 76.3 0/3 2/3 ( H2 ) 18 6,768 128 s $0.271
G003 · App Service 65.4 (59.9–69.4) 0.722 0/5 60.5 0/3 3/3 ( H1 ,H2 ) 11 7,736 231 s $0.238

Rubric items not at full credit, counted over attempts that planned: G001: H6 2/3, H7 2/3, H9 3/3, H12 3/3, H13 3/3, H15 3/3; G002: A4 1/3, B1 2/3, H2 2/3, H4 1/3, H5 1/3, H8 1/3, H9 3/3, H11 1/3; G003: H1 3/3, H2 3/3, H4 1/3, H7 1/3, H8 3/3. Definitions are in Scoring criteria.

ipv4_enabled = false is a literal, and the IAM authentication flag is merged after consumer flags so it cannot be overridden.H1 and H2 (security findings, 3/3).ip_configuration omits H8, 0/3); H9, H13). sku_name, disk_type, availability_type). G003/gemini-3.1-pro-preview-e1/main.tf lines 45–58

resource "azurerm_linux_web_app" "this" {
  name                = var.app_name
  location            = azurerm_resource_group.this.location
  resource_group_name = azurerm_resource_group.this.name
  service_plan_id     = azurerm_service_plan.this.id

  identity {
    type = "SystemAssigned"
  }

  site_config {
    application_stack {
      docker_image_name   = var.container_image
      docker_registry_url = var.container_registry_url

An otherwise complete resource that never sets https_only or the basic-auth publishing flags. The same omission appears on all six App Service web apps and slots this model wrote.

Insecure configuration is scored by the rubric, not the judge, so the security findings move this model's judge grade very little. On Cloud SQL, the attempt with a security finding received the highest judge grade of the three, and the secure attempt the lowest. The composite takes both into account. Read either component alone and you get a misleading picture.

How Claude Fable 5.1 did on iac-bench 2026-10: an overall composite of 86.9, scores per fixture, strengths, weaknesses and the patterns in its OpenTofu modules.

How Claude Fable 5.2 did on iac-bench 2026-10: an overall composite of 93.9, scores per fixture, strengths, weaknesses and the patterns in its OpenTofu modules.

How Claude Haiku 5.5 did on iac-bench 2026-10: an overall composite of 68.0, scores per fixture, strengths, weaknesses and the patterns in its OpenTofu modules.

── more in #ai-research 4 stories · sorted by recency
── more on @gemini 3.1 pro 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/gemini-3-1-pro-on-ia…] indexed:0 read:3min 2026-11-04 · —