Redis 8.0 Lands: New Features and More License Drama

Last week, Redis launched version 8.0 GA, delivering its first new data type in years—vector set—along with unified Redis Stack capabilities (JSON, time series, search, etc.) baked into the core distribution, plus performance improvements. But the bigger headline? Yet another licensing pivot.

After the firestorm of March 2024 (when Redis abandoned the BSD open-source license for RSALv2/SSPLv1), the company has now partially backtracked, adopting a tri-license model with AGPLv3 alongside existing terms. While some celebrate a return to open-source roots, others are skeptical because, let’s face it, Redis’s licensing rollercoaster hasn’t exactly inspired universal trust.

In this post, we’ll break down what Redis 8.0 actually means for developers, the open-source community, and business organizations—the highs (new features), the headaches (license whiplash), and how competition is pushing Redis to evolve.

Licensing Whiplash: Two Steps Back, One Step Forward

The License Change Backfired

When Redis abandoned its BSD license in March 2024 for the controversial SSPL, it achieved its immediate goal—AWS and GCP were forced to create their own fork. But as Redis Inc. CEO Rowan Trollope now admits, this controversial pivot came at an enormous cost. The move didn't just prevent hyperscalers from monetizing Redis-as-a-Service without contributing back—it burned bridges with the community that had nurtured the project for over a decade, while simultaneously alienating Linux distributions committed to open-source principles with Valkey now occupying Redis' former place in major distros.

A (Partial?) Return to Open Source

In November 2024, Salvatore Sanfillipo (aka antirez, the original author of Redis) rejoined Redis Inc. and influenced the company to add AGPLv3 as an additional licensing option for Redis. Announced on May 1, 2025, Redis 8.0’s new tri-license model (RSAL, SSPL, or AGPL) represents a partial course correction. By adding an OSI-approved option, Redis is clearly trying to win back developer goodwill and Linux distribution inclusion. While the inclusion of AGPL signals reconciliation with open-source values, its actual adoption may prove limited—many enterprises avoid AGPL due to its restrictive copyleft requirements, particularly the "use over network" clause that creates compliance headaches.

Redis Licensing Overview (OSS, Enterprise, and Cloud)

Developer Distrust Runs Deep

The community reaction has been...complicated.

Hacker News user @c0l0’s comment captures the prevailing sentiment:

I contributed a minor (but imho still neat :p) improvement to Redis under its original license and personally moved to using Redict when the unexpected license change to SSPL was announced—and I was feeling betrayed as a contributor to a proper FOSS codebase. (Had they switched to AGPL right away, I’d have been perfectly fine with that change from a moral perspective, ftr.)

I have a great deal of respect for antirez and recognize him as a kind and benevolent member of the FOSS community, but no matter what Redis, Inc. announces or does, they have lost my trust for good, and I will continue to use Redis forks for as long as they exist.

The open-source world operates on trust, and Redis spent decades building theirs only to squander it in a single controversial decision. The fallout was immediate and lasting: core contributors permanently migrated to forks, while hyperscalers like AWS and GCP, now backing Valkey, will not likely return to offering new Redis versions.

Redis Forks: When Ideology Became Innovation

What makes this situation particularly uncertain for Redis isn’t just the existence of forks—it’s that projects have evolved from ideological alternatives into legitimate technical competitors. And this isn’t necessarily negative. Competition forces innovation, and the in-memory data store ecosystem ultimately benefits. Consider the progress already made—Valkey has outpaced Redis in certain performance benchmarks as well as optimized its hash table implementation.

This is the paradox of Redis’s licensing strategy: in trying to control commercialization, they’ve unwittingly spawned competitors with cleaner governance and faster innovation cycles.

The Hyperscaler Paradox

There’s undeniable irony here. Redis’s original license change aimed to stop hyperscalers from profiting off their work without fair contribution—but the resulting forks may ultimately hurt Redis the project and the community more. As Hacker News user @echelon notes:

The solution is clear: start your new database with an "equitable source/source available" license from day one. Nobody will complain about a relicense since your license will handle the hyperscalers right off the bat.

Interestingly, this is exactly what Dragonfly is doing—launching with a source-available license (BSL) from the outset, seeking a deliberate balance between open collaboration, permissive use, and sustainable business. The approach works well for us—even with BSL, Dragonfly’s community thrives, and adoption grows rapidly. No license whiplash, no fractured trust. Just a clear path forward for everyone.

Redis 8.0: New Features and Competitive Responses

More Than Just Licensing

When Salvatore rejoined Redis as a developer evangelist, his influence extended beyond the open-source resumption. A more tangible contribution is vector sets—Redis’s first new core data type in years, marking a return to technical innovation after a long focus on enterprise features (i.e., active-active replications). Inspired by sorted sets, vector sets enable high-dimensional similarity search—a nice addition for AI workloads like semantic search.

The Redis Stack Integration

But the new offering does not stop at vector sets. Redis 8.0 also bundles Redis Stack data types and features (i.e., JSON, Search) directly into a single distribution. This consolidation not only simplifies deployment but also strengthens Redis’ position against Valkey in the battle for Linux distribution inclusion. Notably, Redis is only now consolidating Redis Stack features into the core distribution, while Dragonfly implemented these capabilities natively from the start, eliminating fragmentation and add-on complexity.

Performance Improvements

After years of focusing on enterprise features while core development lagged, Redis is now working to catch up with challengers like Dragonfly and forks like Valkey. Redis claims "up to 87% reduction in command latency" and "2x more ops/sec throughput by enabling I/O multithreading" compared to Redis 7.2.5. While we haven’t tested these improvements ourselves yet, it’s exciting to see performance becoming a priority again.

This shows how competition drives progress. The emergence of more performant, more innovative options has pushed Redis to refocus on its technical roots. Whether these changes are catching up fast enough, it’s good for users either way. And it keeps all of us working harder to push in-memory data performance even further.

Redis at the Crossroads

Redis 8.0 marks a significant shift—a return with an open-source license, new feature additions, and clear responses to the changes in the market. By changing their license and improving their product, Redis is trying to win back the community and the distributions.

For developers, this is all good news. After all this drama, the real winners are developers, who now have more features to use and performance gains to access in the in-memory data store landscape. As for the future? With Dragonfly, Redis, and Valkey now established as viable competitors in the in-memory data store space—it’s everyone’s race to win.