Bottom-tier model prices have converged across the US labs and Chinese open weights undercut them, so the decisions that separate vendors now concern who answers for an agent that goes wrong
Frontier price competition has collapsed the cost of capable models to near-parity at the bottom tier, while the real contest has moved to who is accountable when agents go wrong. Governments are responding with incident, safety-proof and reporting demands rather than capability rules. This week Anthropic priced its smallest new model identically to OpenAI's cheapest tier, and a Chinese open-weight rival sits in the same range. With price no longer separating vendors at the entry level, attention shifted to breaches and disclosures. Investigators in Australia and California pressed OpenAI. A White House task force turned Anthropic's incident disclosures into a reporting mandate. Australia outlined plans to make AI companies prove their safety systems work. None of these responses limits what models can do. Each asks who knew what, when, and whether the safeguards hold.
$0.15 DeepSeek V4.1-Flash input price per 1M tokens, cache miss, off-peak
$0.003 DeepSeek V4.1-Flash input price per 1M tokens on a cache hit
$2/$10 GPT-6.1 Sol input and output price per 1M tokens
Signal Pro adds the evidence behind the argument, how this week sat against previous weeks, the policy read for every market and what it changed for planning.
Get Signal Pro →The week's read in your inbox every Monday. Unsubscribe anytime.
A research seat runs Mireto's research on any organization in your plan's market and helps you prepare the conversation. Partner firms deliver the result to their clients in their own name.