Mythical Games to Launch FIFA Rivals Web3 Mobile Game in June

TheCryptoTimes發佈於 2025-04-24更新於 2025-04-24

Web3 gaming studio Mythical Games is gearing up to launch its much-anticipated soccer game, FIFA Rivals, in June. Inspired by FIFA and built for mobile, the game will go live globally alongside this year’s FIFA Club World Cup, following a pre-release in May.

The game FIFA Rivals will be released for both iOS and Android platforms with integrating the concept of blockchain for football enthusiasts. Mythical Games is working on the title in collaboration with the Colombian studio Bacon Games.

This launch follows the studio’s success with “NFL Rivals”, which drew nearly 3 million players by late 2023. Known for strong partnerships, Mythical Games is among the few Web3 studios successfully working with major sports brands like the NFL and FIFA.

FIFA has previously explored Web3 through digital collectibles that offered owners special perks, including a shot at tickets to the 2026 World Cup final.

Mythical Games has built strong momentum in the Web3 space, raising $150 million in a Series C round led by a16z in 2021. The round pushed the company’s valuation to $1.25 billion. Other big-name backers include Cathie Wood’s ARK Invest, Animoca Brands, MoonPay, PROOF, and Stanford Athletics.

It is also developing a new mobile game called Pudgy Party that will be based on the Pudgy Penguins NFT brand and is expected to be released in 2025.

With FIFA Rivals, Mythical Games plans to hit the goalpost by leveraging football enthusiasts and the blockchain technology.

Also  Read: Hamster Kombat GameDev Cipher Code And Combo Card April 23



你可能也喜歡

具身智能“高考”难疯了,人类100分,最强模型12.8

近日,一个名为RoboDojo的新型机器人操作评测基准发布,旨在全面评估具身智能模型的综合能力。该基准由学术机构联盟推出,包含42个仿真任务和18个真实世界任务,覆盖泛化、记忆、精细操作、长程执行和开放语义理解五大核心维度。 评测结果显示,当前最强通用机器人策略在仿真任务中的平均成功率仅为8.80%,在真实世界任务中最好模型的平均成功率也仅有12.8%。相比之下,人类专家在仿真和真实任务中的成功率分别达到76.03%和100%,差距显著。 RoboDojo的仿真任务设计复杂,例如在泛化任务中场景物体可达25个且随机变化,记忆任务需识别曾在传送带出现的物体,精细操作要求完成插管、对齐等高精度动作。其实验室还设立了标准化的真机评测平台(RoboDojo-RealEval),使用多款双臂机器人执行如盖积木、做面包、插充电器等日常任务,并统一硬件、布局与评测流程,确保结果可复现、可比较。 在公开排行榜中,仿真任务榜首Hy-Embodied-0.5-VLA平均成功率仅8.80%,真实任务榜首π0.5成功率12.8%。模型表现不均,尤其在开放语义任务上,最强模型成功率仅约1.67%,表明现有系统在理解新指令并可靠执行方面仍很薄弱。 为支撑评测,项目同时提供了异构并行仿真工具和统一策略接入层XPolicyLab,帮助研究者高效集成和比较不同模型。团队表示,RoboDojo如同“具身智能的珠峰”,旨在通过标准化、多维度的评测推动领域发展,未来还将扩展灵巧操作、移动操作等更多评测方向。

marsbit3 分鐘前

具身智能“高考”难疯了,人类100分,最强模型12.8

marsbit3 分鐘前

交易

現貨
活动图片