Opus 5.5 Shows No Performance Drop on Functional Languages in 202-Task Benchmark

·2026.11.10 23:18

Key point

A new benchmark of 202 real-world tasks across 13 languages finds Opus 5.5 maintains performance on functional languages while open-weight models like DeepSeek, GLM, and Qwen fall behind.

Details

Benchmark evaluates 202 tasks from real issues and PRs across 13 languages, including Elixir, OCaml, Erlang, Clojure, Kotlin, and Swift.
On functional languages (Elixir, OCaml, Erlang), the gap between closed and open-weight models is wider; Opus 5.5 shows no drop in performance, while DeepSeek, GLM, and Qwen fall further behind.
Each task uses a dedicated Docker image, and models were run three times per task to ensure fairness.
Full leaderboard, paper, and model trajectories are available on longtail-swe.com, Hugging Face, and Harbor.

software-engineering anthropic coding llm-benchmark opus-5.5

Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.