{"rewrite":{"id":"r_dc8eda48ceb39b55aa94dc20","clusterId":"c_a94f5eefe484d62cf7493b20","slug":"trans-n-ships-swama-2-5-1-with-openai-decisions-api-support","model":"deepseek-v4-1-flash","headline":"Trans-N Ships Swama 2.5.1 With OpenAI Decisions API Support","summary":"Trans-N released version 2.5.1 of Swama, an open-source platform for running large language models locally on Apple silicon Macs. The build supports the OpenAI Decisions API format, so classification, yes/no decisions, and graded evaluations on text and images run on the machine without a cloud inference call. In internal testing on a Mac Studio (M3 Ultra), 30 inquiries classified in 4.5 seconds and 30 product photos assessed in 10.5 seconds.","whyItMatters":"Trans-N is positioning local decision processing, not text generation, as the practical use case, and its own test numbers put a locally hosted Qwen3.5-35B model ahead of a cloud API call on time while landing in the same accuracy range.","webCardHtml":"\u003cp\u003eThe release centers on a decision endpoint, \u0026#34;/v1/decisions\u0026#34;, built for routing inquiries to a department or judging whether a product photo shows damage, rather than for producing long text. Applications call it through the API and receive the result with a confidence value from the model. Trans-N notes limits on supported input formats and image counts, and states that confidence does not guarantee a correct answer; the supported scope is listed in the GitHub documentation.\u003c/p\u003e\u003cp\u003eThe comparison numbers come with stated caveats. Both sides started at once and processed one item at a time, timed as the median of three runs through a common local proxy. Trans-N says results vary by model, input, hardware, and network conditions, and that the measurement does not demonstrate superiority in speed or accuracy across all environments.\u003c/p\u003e","blueskyPost":"Trans-N released Swama 2.5.1, an open-source platform for running LLMs locally on Apple silicon Macs. New: support for the OpenAI Decisions API format, so classification and image judgment run on the machine. Internal test: 30 inquiries in 4.5 seconds.","twitterPost":"Trans-N released Swama 2.5.1, an open-source platform for local LLMs on Apple silicon Macs. It adds OpenAI Decisions API format support for text and image decisions. Internal test on a Mac Studio: 30 inquiries classified in 4.5 seconds.","threadsPost":null,"newsletterBlurb":"Trans-N shipped Swama 2.5.1, an open-source platform for running large language models locally on Apple silicon Macs. The update adds support for decisions in the OpenAI Decisions API format, covering text and image classification and graded evaluation on the machine. In internal testing, a Mac Studio running a local Qwen3.5-35B model classified 30 inquiries in 4.5 seconds.","attributionJson":"[{\"source\":\"ASCII.jp\",\"url\":\"https://ascii.jp/elem/000/004/440/4440922/?rss\",\"title\":\"Trans-N Releases Local AI Execution Platform 'Swama' 2.5.1 - Text and Image Decision Processing on a Single Mac\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":4671,"outputTokens":666,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1791373932,"createdAt":"2026-10-07T11:47:10.000Z","publishedAt":"2026-10-07T11:51:07.000Z","updatedAt":"2026-10-07T11:51:07.000Z"},"cluster":{"id":"c_a94f5eefe484d62cf7493b20","canonicalTitle":"Trans-N、ローカルAI実行基盤「Swama」2.5.1を公開　文章・画像の判定処理をMac 1台で","representativeArticleId":"a_89b6b22d5838c3ebba514698","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[\"Swama\"],\"studios\":[\"Trans-N\"],\"people\":[],\"type\":\"release\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-10-07T09:51:31.000Z","lastSeenAt":"2026-10-07T09:51:31.000Z","updatedAt":"2026-10-07T11:51:06.000Z"},"attribution":[{"source":"ASCII.jp","url":"https://ascii.jp/elem/000/004/440/4440922/?rss","title":"Trans-N、ローカルAI実行基盤「Swama」2.5.1を公開　文章・画像の判定処理をMac 1台で"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":["Swama"],"studios":["Trans-N"],"people":[],"type":"release","domain":"other","is_roundup":false},"keyFacts":null}