Observed Signal · Oct 9, 2026 · Technical Release · Source: OpenAI Blog · Impact: 4/5 · Sentiment: Positive
Asana Cuts Browser Agent Costs 76x with GPT-6.1 Sol
Asana leveraged OpenAI's GPT-6 Astra in Codex to optimize its browser agent's workflow on GPT-6.1 Sol, achieving a 76x cost reduction and 5x speed improvement. The study, conducted by StackAI CTO Frank Hidalgo, tested four models and implemented caching and screenshot policies that reduced estimated model costs from $36.21 to $0.47 per run. The optimization allowed Asana to offer faster, more capable models to customers while lowering operational costs. Asana has released the changes to StackAI and plans to incorporate similar testing into platform evaluations.
Major AI model optimization showcases significant cost and efficiency gains for enterprise automation, influencing AdTech operations that rely on AI agents.
Track Asana Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Asana optimized its browser agent on GPT-6.1 Sol, reducing model costs from $36.21 to $0.47 per run (76x cheaper) and speeding up runs from 22.5 minutes to 4 minutes (5x faster).
- The optimization used GPT-6 Astra in Codex to automate experiments, cutting research time from an estimated 1-2 months to about a week.
- The optimized workflow retained browsing history of up to 480,000 characters and used a caching policy that reduced input costs by 89% (5% of uncached price).
- Asana has released the browser navigation changes to StackAI and plans to integrate similar testing into platform evaluations.
- The study tested four models including GPT-6.1 Sol and three others; the optimized setup on GPT-6.1 Sol completed all runs with correct answers.
Connected Companies & Entities
2 Entities mapped“Asana helps customers automate work across business applications through StackAI, a platform it acquired.”
“The article describes Asana's use of GPT-6 Astra in Codex and GPT-6.1 Sol, both OpenAI models.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Parallel cuts research time and cost in half with GPT-6 Astra
OpenAI's case study highlights how Parallel, a developer of AI agent infrastructure, achieved significant efficiency gains using OpenAI's GPT-6 Astra model. In a test involving the compilation of labor-market statistics data, GPT-6 Astra completed the research in half the time and with roughly 50% cost reduction compared to prior models. The model demonstrated more focused searches, requiring fewer steps to reach high-quality results, and enabled effective delegation to sub-agents for parallel task execution. This has practical implications for AI-driven research automation, offering reduced latency and operational costs for complex web-based knowledge work.
OpenAI introduces GPT-6.1 Sol, cheaper near-Astra model
This week's AI/ML news highlights significant model releases and a trend toward specialized, cost-efficient AI. OpenAI introduced GPT-6.1 Sol, a cost-efficient model for complex coding tasks, scoring 52 on the Artificial Analysis Intelligence Index—one point below GPT-6 Astra—with a 1.05M-token context window and API pricing at $2/M input and $10/M output tokens. Sol excels in agentic coding and computer use, outperforming GPT-6 Sol on benchmarks, and is available in ChatGPT Work and Codex. Notably, OpenAI cancelled GPT-6.1 Astra due to safety concerns. Emerson also launched Claude Sonnet 5.5, claiming faster output and lower costs, while Cloudflare released Clef decision models, NVIDIA unveiled Kumo Tabular, and Google announced Gemini 4 Argon. The overarching theme is the shift toward heterogeneous AI architectures.
Basis Doubles Tax Workbook Speed with GPT-6 Astra
Basis, a company specializing in AI agents for accountants, has reported that using OpenAI's GPT-6 Astra model enables its agents to complete a complex 50-tab tax workbook twice as fast as with the previous GPT-5.6 Sol model. The improved performance is attributed to GPT-6 Astra's enhanced understanding of user intent, better initial decision-making, and adaptive reasoning effort. Basis observed about a 20% improvement in internal evaluation scores, driven by the model's ability to infer expectations with fewer explicit instructions. Additionally, GPT-6 Astra can adjust computation during tasks while keeping cache intact, reducing costs and response times for long-running operations.
Receive all future signals for Asana & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.