PulseAugur
EN
LIVE 17:34:37

Qwen3.8-27B lags Ornith 1.5 in Mac speed tests

By PulseAugur Editorial · [1 sources] ·

A recent comparison on a Mac Studio M4 Max evaluated the performance of Qwen3.8-27B and Ornith 1.5 models. Ornith 1.5 demonstrated significantly faster first-token generation, taking 0.49 seconds compared to Qwen3.8-27B's 3.46 seconds for an 800-token prompt. While Qwen3.8-27B achieved a respectable output speed of 55 tokens per second, Ornith 1.5 was faster at 90 tokens per second. In code review tests, both models performed similarly when reasoning was disabled, but Qwen3.8-27B struggled when reasoning was enabled, producing fewer correct bug identifications and often exceeding output limits. AI

IMPACT Highlights performance differences in local LLM inference, impacting user choice for Mac-based AI tasks.

RANK_REASON Comparison of two specific LLMs on local hardware using specific tools.

Read on dev.to — LLM tag →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Qwen3.8-27B lags Ornith 1.5 in Mac speed tests

How we ranked this

Signal score
39 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
Comparison of two specific LLMs on local hardware using specific tools.
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. dev.to — LLM tag TIER_1 English(EN) · Javier Rodeiro Rodríguez ·

    Qwen3.8 vs Ornith 1.5 on a Mac: speed and code review

    <p>On a Mac Studio M4 Max with 64 GB, Qwen3.8-27B with MTP does not replace the 35B mixture-of-experts model I already run. Measured on 12 October 2026 under oMLX 0.7.0, Qwen generated 55 tokens per second against 90, and took 3.46 seconds to produce its first token on an 800-tok…

RELATED ENTITIES

RELATED TOPICS