--- title: [AINews] not much happened today | daily.dev description: A daily roundup covering AI news from Twitter and Reddit: OpenAI's GPT-6.1 Sol launches at a steep price cut while claiming to beat prior models and rivals on... canonical: https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: [AINews] not much happened today | daily.dev og:description: A daily roundup covering AI news from Twitter and Reddit: OpenAI's GPT-6.1 Sol launches at a steep price cut while claiming to beat prior models and rivals on... og:url: https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui og:image: https://api.daily.dev/og/posts/KMwwJGwuI.png og:image:alt: [AINews] not much happened today og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # [AINews] not much happened today **[Latent Space](https://daily.dev/sources/latentspace)** · 29 min read · 0 upvotes · 0 comments ## Summary A daily roundup covering AI news from Twitter and Reddit: OpenAI's GPT-6.1 Sol launches at a steep price cut while claiming to beat prior models and rivals on coding/agent benchmarks; Anthropic's Sonnet 5.5 and Opus 5.5 dominate Agent Arena rankings; rumors swirl about Claude Fable 5.5 and a delayed GPT-6 Astra. Developer tooling updates include llama.cpp adding decision-model inference, Pi 1.0 shipping with native MCP support, DeepSeek Harness desktop builds, and Cloudflare's open-weights Clef decision model. Research highlights cover multi-harness RL training, long-horizon agent control, AI-assisted math research, and new research-taste benchmarks. Hardware coverage touches Huawei's Ascend 950 chip analysis, Prime Intellect's NVFP4 inference optimizations, and NVIDIA memory architecture claims. Reddit threads debate whether local 27B models are closing the gap with frontier models, report user-perceived Opus 5.5 quality regressions, and discuss Gemini 4 Argon's rocky access rollout alongside new AI video generation models like Griffin and MiniMax orbit LoRAs. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does GPT-6.1 Sol's pricing compare to GPT-6 Astra? GPT-6.1 Sol is priced at $2 per million input tokens and $10 per million output tokens, compared with $10/$50 for Astra. Despite the much lower price, Sol reportedly beats GPT-6 Sol by 6.4 points on DeepSWE v1.1 and beats Opus 5.5 by 2.2 points on AutomationBench, positioning it as a cheaper, faster alternative for coding and automation tasks. _Developers weighing model costs against coding benchmark performance can track pricing shifts like this on daily.dev._ ### What new decision-model support did llama.cpp add for local inference? Llama.cpp added a /v1/systemone endpoint enabling local 'Jev-style' decision-model inference, letting developers run classifier-like controller models locally, for example via the command llama serve -hf ggml-org/Kev-4B-GGUF. This complements releases like Cloudflare's Clef decision model, post-trained from Qwen3.8-27B, and a smaller clef-flash variant post-trained from Qwen3.5-9B. _Anyone experimenting with local agent control loops can follow llama.cpp tooling updates like this on daily.dev._ ### Are users reporting that Claude Opus 5.5 got worse after launch? Yes, multiple users report a perceived quality regression in Claude Opus 5.5 within days of launch, describing verbose preambles, duplicated code, and token-hungry 'slopcode' behavior resembling the prior Opus 5 model. One user's token usage jumped from roughly 70% to 90% in about an hour, and third-party sentiment tracking on modelsentiment.com showed scores dropping from 71-73/100 to 55-58/100 over a few days, though no controlled benchmark confirms an actual backend change. _Developers relying on Claude Opus for coding workflows can watch for regression reports like this on daily.dev._ ## Similar posts on daily.dev - [\[AINews\] The Future of Latent Space](https://daily.dev/posts/ainews-the-future-of-latent-space-gzusgggls) · Latent Space · 0 upvotes · 0 comments - [OpenAI launches GPT-6 Sol and Luna, boasting lower cost and fewer mistakes](https://daily.dev/posts/openai-launches-gpt-6-sol-and-luna-boasting-lower-cost-and-fewer-mistakes-7fzpvatba) · TechCrunch · 0 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents), [#claude](https://daily.dev/tags/claude), [#gpt](https://daily.dev/tags/gpt), [#llama-cpp](https://daily.dev/tags/llama-cpp) [View this post on daily.dev](https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"[AINews] not much happened today","url":"https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui"},"datePublished":"2026-10-03T08:49:44.453Z","dateModified":"2026-10-03T08:51:30.107Z","description":"A daily roundup covering AI news from Twitter and Reddit: OpenAI's GPT-6.1 Sol launches at a steep price cut while claiming to beat prior models and rivals on...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4048deeae8221ed153906fd42dea8bf2?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4048deeae8221ed153906fd42dea8bf2?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Latent Space","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Latent Space","logo":"https://media.daily.dev/image/upload/s--DnJ9laFj--/f_auto,q_auto/v1780213196/logos/latentspace?_a=BAMAMiWQ0","url":"https://daily.dev/sources/latentspace"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-agents,claude,gpt,llama-cpp","timeRequired":"PT29M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Latent Space","item":"https://daily.dev/sources/latentspace"},{"@type":"ListItem","position":3,"name":"[AINews] not much happened today"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/ainews-not-much-happened-today-kmwwjgwui#faq","mainEntity":[{"@type":"Question","name":"How does GPT-6.1 Sol's pricing compare to GPT-6 Astra?","acceptedAnswer":{"@type":"Answer","text":"GPT-6.1 Sol is priced at $2 per million input tokens and $10 per million output tokens, compared with $10/$50 for Astra. Despite the much lower price, Sol reportedly beats GPT-6 Sol by 6.4 points on DeepSWE v1.1 and beats Opus 5.5 by 2.2 points on AutomationBench, positioning it as a cheaper, faster alternative for coding and automation tasks. Developers weighing model costs against coding benchmark performance can track pricing shifts like this on daily.dev."}},{"@type":"Question","name":"What new decision-model support did llama.cpp add for local inference?","acceptedAnswer":{"@type":"Answer","text":"Llama.cpp added a /v1/systemone endpoint enabling local 'Jev-style' decision-model inference, letting developers run classifier-like controller models locally, for example via the command llama serve -hf ggml-org/Kev-4B-GGUF. This complements releases like Cloudflare's Clef decision model, post-trained from Qwen3.8-27B, and a smaller clef-flash variant post-trained from Qwen3.5-9B. Anyone experimenting with local agent control loops can follow llama.cpp tooling updates like this on daily.dev."}},{"@type":"Question","name":"Are users reporting that Claude Opus 5.5 got worse after launch?","acceptedAnswer":{"@type":"Answer","text":"Yes, multiple users report a perceived quality regression in Claude Opus 5.5 within days of launch, describing verbose preambles, duplicated code, and token-hungry 'slopcode' behavior resembling the prior Opus 5 model. One user's token usage jumped from roughly 70% to 90% in about an hour, and third-party sentiment tracking on modelsentiment.com showed scores dropping from 71-73/100 to 55-58/100 over a few days, though no controlled benchmark confirms an actual backend change. Developers relying on Claude Opus for coding workflows can watch for regression reports like this on daily.dev."}}]} ```