Two release trackers show no new AI models over the 17–18 October weekend beyond DeepSeek's V4.1-Flash-1017, and Mistral's Large 4 weights are still “Coming soon”

The weekend of 17 and 18 October brought no new AI model releases beyond one the Switchboard ran yesterday. DeepSeek's V4.1-Flash-1017, which its news page dates 17 October, is the newest entry on LLM Gateway's release timeline, which says it was updated on 17 October. The timeline's next entries are Grok Voice 3.0 (14 October) and Muse Wonder 1.0 (12 October), and nothing on it is dated 18 October. A second tracker, Evertune's, runs to 15 October, with nothing later than Meta's Muse Spark Open.

One promised release is still pending. Mistral's documentation for Large 4, which it released as a preview on 6 October, still lists “Open weights: Coming soon”, with the status “Public Preview”. When the Switchboard reported the preview on 7 October, Mistral said the weights would come out by the end of the month; the page gives no date.

On the leaderboards, Artificial Analysis's speech-to-speech page puts SpaceXAI's Grok Voice 3.0 first with a score of 86.9, ahead of GPT-Live-1 (Astra, medium) at 83.2. Grok Voice 3.0 itself ran as a story on 15 October; the ranking is an update.

A quiet weekend is a finding in its own right: the week's model news is what the Switchboard has already run, and the next thing to watch on the open-weights side is whether Mistral's Large 4 weights arrive before October ends.

• Confirmed LLM Gateway's timeline, updated 17 October, lists DeepSeek V4.1 Flash 1017 (17 October) as its newest release, with nothing dated 18 October. LLM Gateway, model timeline
• Confirmed Evertune's model tracker runs as of 15 October, with nothing later than Muse Spark Open. Evertune, AI model tracker
• Confirmed Mistral's documentation for Large 4 lists “Open weights: Coming soon”. Mistral documentation
• Confirmed Grok Voice 3.0 leads Artificial Analysis's speech-to-speech ranking with 86.9. Artificial Analysis