Hassan Khan
In an effort to cut user costs, OpenAI has introduced Flex manufacturing, a new API option that trades reduced rates for longer response times and sporadic resource outages.
Flex, which targets low-priority activities like enhancing data and async work, is available for o3 and o4-mini models and is currently in beta.
O3 now costs US$5 per million incoming tokens and US$20 per million output tokens, which is half the usual prices for Flex.
In contrast to earlier rates of US$1.10 and US$4.40, prices for the o4-mini model decrease to US$0.55 per million input tokens and US$2.20 per million output tokens.
The debut coincides with the introduction of less expensive AI models, such the Gemini 2.5 Flash, by competitors like Google.
Read More: OpenAI releases its ‘smartest’ models yet
Flexible consumption models drive AI pricing
The launch of “Flex processing” by OpenAI at precisely half the cost of conventional solutions marks a substantial advancement in AI pricing tactics that is consistent with general market trends.
Based on workload requirements, clients may select between premium (on-demand) and economy (spot/reserved) choices in this new tier, which has intentionally slower performance and “occasional resource unavailability,” creating a price dynamic akin to what revolutionized cloud computing.
Read More: OpenAI Launches Powerful New Reasoning AI Models with Bonus Features
The strategy makes financial sense for both parties: cost-conscious developers may access robust models at significantly cheaper prices, while OpenAI can better utilize its computational resources by shifting non-time-sensitive workloads to times when demand is lower.
With Google’s concurrent release of Gemini 2.5 Flash positioned as a more cost-effective substitute that “matches or bests DeepSeek’s R1 efficiency at a cheaper input token cost,” this price breakthrough comes amid escalating rivalry.
When creating AI monetization methods, the industry seems to be shifting toward more complicated, usage-based pricing structures that follow the advice of pricing experts, taking into account “complexity, user value, and market demand”Â
The role of identity verification in responsible AI
A significant change in the way AI businesses handle governance and security issues is shown by OpenAI’s new need for verification of identity for higher-tier API users.
OpenAI claims that this authentication requirement is intended to “stop bad actors from breaking its usage policies.” It comes at a time when worries about fraud and misuse facilitated by AI are on the rise.
The timing is significant because generative AI technologies are being used to produce deep fakes and fake identities that may evade conventional verification methods, as security experts have recently revealed.
In a real-world instance cited by risk intelligence company LSEG, people have effectively created fake identities using generative AI to trick security systems.
This leads to a difficult situation: although OpenAI is putting ID verification in place to stop model abuse, the same sophisticated AI models are also posing a threat to established identity verification practices in other sectors.
The action exemplifies the difficult balancing act AI providers must perform: providing developers with access to robust models while putting in place adequate protections against possible damage.
| Date | Development/Update | Description |
|---|---|---|
| 17-Apr-2025 | OpenAI considers social media platform for AI-generated images | OpenAI developing AI-focused social media platform New features enhance content sharing capabilities |
| 17-Apr-2025 | OpenAI releases its ‘smartest’ models yet | OpenAI launches o3 and o4-mini models New models improve reasoning and math capabilities |
| 17-Apr-2025 | OpenAI reportedly in talks to acquire AI coding firm for $3B | OpenAI in talks to acquire Windsurf for $3B Deal boosts OpenAI’s AI coding tools market position |
| 16-Apr-2025 | OpenAI said to build social media platform | OpenAI developing a new social media platform Potential competition with X and Meta in AI space |
| 15-Apr-2025 | OpenAI introduces GPT-4.1 models with 1m tokens for automation | OpenAI launches GPT-4.1, mini, and nano models New models focus on coding, efficiency, and cost |
| 14-Apr-2025 | OpenAI mulls ID verification for access to future AI models | OpenAI plans ID verification for API access One org per ID verified every 90 days |
Conclusion
OpenAI’s Flex API marks a turning point in affordable AI deployment. It gives developers access to powerful models at half the cost, reshaping how AI resources are allocated. Combined with stronger user verification, OpenAI is not only reducing financial barriers but also reinforcing responsible AI use. As competition heats up, flexible pricing and trust-based access could define the next phase in the AI market.
