Most AI platforms provide users the choice to opt-out of sharing usage data with developers to enhance future iterations. However, Meta has introduced a unique approach by attaching a price to this data access.
For its new Muse Spark model, designed to operate coding and other autonomous agents, Meta is offering substantial discounts. Users can receive an average reduction of approximately 95% if they "contribute" to future model development by sharing their prompts and model outputs. Under the standard agreement, one million input tokens cost $1.25, but under the contributor model, they drop to just $0.10. Similarly, the standard price for one million output tokens is $4.25, while the contributor tier reduces this cost to $0.20.
Meta has faced challenges in acquiring training data in the past; an initiative to track employee computer usage earlier this year faced internal criticism and was paused in June. The company declined to comment on its new pricing strategy regarding TechCrunch.
Usage data is crucial for enhancing agentic tools. Mario Zechner, the developer behind the open-source Pi harness, explained to TechCrunch last month, "The reason we saw a big jump in [coding agent] capabilities between April 2025 and October 2025 was that Claude Code, by default, would store all your coding agent sessions and use them for reinforcement learning training."
Despite the growing imperative for companies to deploy agentic tools beyond software engineering, their ability to evaluate and improve these tools is often hindered by the lack of digital traces in complex professional workflows.
Arvind Narayanan, a computer science professor at Princeton, pointed out evidence suggesting large corporations prefer to retain their data. He noted on social media, "They stick with token-billed Enterprise plans even though the subscription-based consumer plans like Claude Max and ChatGPT Pro are discounted by 10x-20x or even more! (The main difference between the plans is data retention + enterprise IT governance)."
Recognizing these dynamics, Meta is offering explicit compensation to obtain this information. The contributor tier is intended to "lower the barrier to entry for prototyping, testing integrations, and scaling experiments where training on your data is acceptable."
Narayanan suggested that this framework could encourage large enterprises to be more discerning about what constitutes proprietary data versus information that can be shared with model providers. This move also fits into the broader landscape of price competition among leading AI labs, following Anthropic's recent release of the Fable and Mythos models with lower cached token costs and OpenAI's major price cuts in late July.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.