{"slug": "pair-a-frontier-model-with-a-cheap-sidekick-to-cut-coding-costs", "title": "Pair a frontier model with a cheap sidekick to cut coding costs", "summary": "Artificial Analysis benchmarked Cognition's Devin Fusion, which pairs a frontier model with the cheaper SWE-2 model as a sidekick, finding the GPT-6 Astra pairing costs 43 percent less and runs 31 percent faster than a Claude Fable 5.1 pairing while scoring close behind it. The same roundup reported OpenAI moved its biological reasoning model GPT-Rosalind from research preview to general availability across the API, Codex, and ChatGPT Enterprise with new Life Sciences plugins, and that DeepSeek V4.1 Flash processed 1 trillion tokens in its first 24 hours on OpenRouter, with 90 percent served from cache at about $0.006 per million tokens.", "body_md": "Artificial Analysis benchmarked Cognition's Devin Fusion, which pairs a frontier model with the cheaper SWE-2 model as a sidekick. The GPT-6 Astra pairing costs 43 percent less and runs 31 percent faster than a Claude Fable 5.1 pairing while scoring close behind it.\nRead: Artificial Analysis benchmarked Cognition's Devin Fusion, which pairs a frontier model with the cheaper SWE-2 model as a sidekick. The GPT-6 Astra pairing costs 43 percent less and runs 31 percent faster than a Claude Fable 5.1 pairing while scoring close behind it.\nRead: OpenAI moved GPT-Rosalind, its biological reasoning model, out of research preview into general availability across the API, Codex, and ChatGPT Enterprise, adding Life Sciences plugins for genomic and protein structure work.\nRead: DeepSeek V4.1 Flash processed 1 trillion tokens in its first 24 hours on OpenRouter, on pace for the largest 48 hour paid model launch yet, with 90 percent of tokens served from cache at about $0.006 per million tokens.\nRead: Sakana AI launched Fugu Max and Fugu Ultra v2, orchestration systems that route tasks across a large pool of open and specialized models, including NVIDIA Nemotron, claiming benchmark wins over Opus 5 without relying on any single closed model.\nRead: Simon Willison and Alex Garcia shipped two Datasette security patches after auditing the codebase with Claude Fable 5.1, GPT-5.6, and GPT-6 Astra, then split verification work so one person wrote the failing test and the other implemented each fix.\nRead: Qwen3.8-27B is now served on Cerebras hardware for fast inference, with Artificial Analysis scoring it near GPT-5.6, DeepSeek V4 Pro, and Claude Sonnet 4.6, though early testers report it underperforms on coding tasks.", "url": "https://wpnews.pro/news/pair-a-frontier-model-with-a-cheap-sidekick-to-cut-coding-costs", "canonical_source": "https://www.vibeleaderboard.ai/intel/brief/2026-09-12", "published_at": "2026-09-12 11:11:14+00:00", "updated_at": "2026-09-12 12:10:58.683371+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products", "ai-tools", "ai-infrastructure"], "entities": ["Artificial Analysis", "Cognition", "Devin Fusion", "SWE-2", "GPT-6 Astra", "Claude Fable 5.1", "OpenAI", "GPT-Rosalind"], "alternates": {"html": "https://wpnews.pro/news/pair-a-frontier-model-with-a-cheap-sidekick-to-cut-coding-costs", "markdown": "https://wpnews.pro/news/pair-a-frontier-model-with-a-cheap-sidekick-to-cut-coding-costs.md", "text": "https://wpnews.pro/news/pair-a-frontier-model-with-a-cheap-sidekick-to-cut-coding-costs.txt", "jsonld": "https://wpnews.pro/news/pair-a-frontier-model-with-a-cheap-sidekick-to-cut-coding-costs.jsonld"}}