{"slug": "kimi-k3-on-iac-bench-2026-10", "title": "Kimi K3 on iac-bench 2026-10", "summary": "Moonshot AI's Kimi K3 scored an overall composite of 82.4 (95% CI 57.3–84.7) on iac-bench 2026-10, ranking 5th of 10 models and making it the highest-scoring open-weight model in the release, according to the benchmark results published for the 2026-10-22 generation and 2026-10-23 judge runs. Across 9 attempts (3 fixtures × 3 samples) Kimi K3 failed 1 gate and produced 4 insecure attempts, all four security findings tracing to a secure default that a consumer can override through an input — notably a Cloud SQL ssl_mode validation that defaults to ENCRYPTED_ONLY but also accepts ALLOW_UNENCRYPTED_AND_ENCRYPTED in all three attempts. The model, accessed via OpenRouter at a list price of $0.24 per million input tokens and $11.94 per million output tokens, cost an estimated $3.96 for all 9 attempts, with fixture medians of 81.1 on S3, 79.2 on Cloud SQL and 86.9 on App Service.", "body_md": "Upcoming Workshop: Citizen Developer Agentic Software Factory on AWS 11.4.26 @ 12PM PST\n\n[Register Here!](https://streamyard.com/watch/umMggUFkVMNu)\n\nHow Kimi K3 did on iac-bench 2026-10: an overall composite of 82.4, scores per fixture, strengths, weaknesses and the patterns in its OpenTofu modules.\n\n| Developer | Moonshot AI | \n| API identifier | `moonshotai/kimi-k3` | \n| Access | OpenRouter | \n| Weights | Open | \n| List price (USD / M tokens, in / out) | 0.24 / 11.94 (OpenRouter, 2026-10-23) | \n| Evaluated | 2026-10-22 (generation) · 2026-10-23 (judge) | \n| Attempts | 9 (3 fixtures × 3 samples) | \n| **Overall composite** | **82.4** (95% CI 57.3–84.7) | \n| Rank | 5 of 10 (95% rank interval 5–9) | \n| Mechanical only / judge only | 80.6 (rank 5) / 86.0 (rank 6) | \n| Failed gates · insecure attempts | 1/9 · 4/9 | \n| Estimated cost, all 9 attempts | $3.96 | \n\nKimi K3 is the highest-scoring open-weight model in the release, placing in the second group just behind the leaders. Its App Service median of 87 is close to the top of the field. Its modules are heavily validated, put identity first, and split into conventional files. Its interfaces are large.\n\nAll four of its security findings come from the same habit: a secure default that a\nconsumer can override through an input. On Cloud SQL, `ssl_mode` defaults to\n`ENCRYPTED_ONLY`, but the validation also accepts `ALLOW_UNENCRYPTED_AND_ENCRYPTED`. This\nhappens in all three attempts, and the variable's description itself says the setting\nshould not be weakened.\n\n| Fixture | Composite median (range) | Rubric median | Expertise | Judge median | ✗ gate | ⚠ security | Median inputs | Median output tokens | Median time | Median cost | \n|---|---|---|---|---|---|---|---|---|---|---|\n| G001 · S3 | 81.1 (67.5–85.6) | 0.767 | 4/7 | 92.1 | 0/3 | 1/3 ( `H1` ) | 11 | 21,595 | 317 s | $0.263 | \n| G002 · Cloud SQL | 79.2 (78.4–80.0) | 0.793 | 6/9 | 81.6 | 0/3 | 3/3 ( `H2` ) | 22 | 31,557 | 494 s | $0.409 | \n| G003 · App Service | 86.9 (25.3–89.4) | 0.889 | 4/5 | 84.2 | 1/3 | 0/3 | 27 | 48,604 | 853 s | $0.624 | \n\nRubric items not at full credit, counted over attempts that planned: G001: `A5` 3/3, `B1` 2/3, `H1` 1/3, `H9` 1/3, `H12` 3/3, `H13` 3/3, `H15` 3/3; G002: `B1` 3/3, `H2` 3/3, `H11` 3/3; G003: `A7` 2/2, 1 attempt failed the gate. Definitions are in [Scoring criteria](https://www.massdriver.cloud/blogs/iac-bench-2026-10-scoring-criteria).\n\n`H2` (security finding, 3/3).`ssl_mode` validation admits\nunencrypted connections (excerpt below). The judge scored `H1` (security finding, 1/3).`allow_public_read`, which turns\noff all four public-access-block settings and grants public `s3:GetObject`.` health_check_path` without\n`health_check_eviction_time_in_min`, which the provider requires together. The production\napp sets both.`deletion_policy` (`H11`, 3/3);` H13`, 3/3);` B1` lost in 5 of the 8 attempts that planned.`minimal.tfvars`.` H15`, 3/3).\n`G002/kimi-k3-e1/variables.tf` lines 200–209\n\n```\nvariable \"ssl_mode\" {\n  description = \"TLS enforcement for connections. The organizational standard is ENCRYPTED_ONLY; TRUSTED_CLIENT_CERTIFICATE_REQUIRED additionally demands mutual TLS. Do not weaken this without a security review.\"\n  type        = string\n  default     = \"ENCRYPTED_ONLY\"\n\n  validation {\n    condition     = contains([\"ALLOW_UNENCRYPTED_AND_ENCRYPTED\", \"ENCRYPTED_ONLY\", \"TRUSTED_CLIENT_CERTIFICATE_REQUIRED\"], var.ssl_mode)\n    error_message = \"ssl_mode must be one of ALLOW_UNENCRYPTED_AND_ENCRYPTED, ENCRYPTED_ONLY or TRUSTED_CLIENT_CERTIFICATE_REQUIRED.\"\n  }\n}\n```\n\nThe description states the standard and warns against weakening it, yet the validation accepts the unencrypted mode. That one list entry cost `H2` in all three Cloud SQL attempts.\n\nHow 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.\n\nHow 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.\n\nHow Claude Opus 5.5 did on iac-bench 2026-10: an overall composite of 88.9, scores per fixture, strengths, weaknesses and the patterns in its OpenTofu modules.", "url": "https://wpnews.pro/news/kimi-k3-on-iac-bench-2026-10", "canonical_source": "https://www.massdriver.cloud/blogs/iac-bench-2026-10-kimi-k3", "published_at": "2026-11-04 00:00:00+00:00", "updated_at": "2026-11-04 20:21:44.076566+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "ai-safety", "ai-tools"], "entities": ["Moonshot AI", "Kimi K3", "iac-bench 2026-10", "OpenRouter", "Claude Fable 5.1", "Claude Haiku 5.5", "OpenTofu", "Cloud SQL"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/kimi-k3-on-iac-bench-2026-10", "markdown": "https://wpnews.pro/news/kimi-k3-on-iac-bench-2026-10.md", "text": "https://wpnews.pro/news/kimi-k3-on-iac-bench-2026-10.txt", "jsonld": "https://wpnews.pro/news/kimi-k3-on-iac-bench-2026-10.jsonld"}}