Three companies, OpenAI, Perplexity, and Microsoft, have recently launched decision-model APIs, sparking a price war with Perplexity offering the lowest cost at $0.02 per million input tokens. OpenAI charges $0.10, while Microsoft is priced at $0.042. These models return structured probabilities rather than prose and have no output charges. Perplexity's lower price is attributed to its smaller, open-source model fine-tuned from Qwen3.8-27B, which allows for self-hosting, unlike OpenAI's closed model. Independent benchmark data is scarce, with most figures coming from company claims, making direct comparisons challenging. AI
IMPACT This price war could accelerate adoption of structured decision-making AI, especially for cost-sensitive applications.
RANK_REASON Multiple major companies launching competing products in a new category triggers significant industry news. [lever_c_demoted from significant: ic=1 ai=1.0]
- Decision index weighting method within habitual domains and its application
- Hugging Face
- Jevíčko
- Microsoft
- OpenAI
- Perplexity
- Qwen3.8-27B
- RAGTruth: a hallucination corpus for developing trustworthy retrieval-augmented language models
- SAS Institute
- WinoGrande
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →