Gradient's banner

Gradient

@Gradient_HQ • 685,698 subscribers

Advancing open intelligence through efficient, accessible infrastructure for AI inference and post-training.

Shorts

DeepSeek V4.1 Flash (1017) solved 8/17 Zero2Repo tasks with dsh, at $3.18 in reported API spend across the full suite. OpenCode scored 7/17 and Claude Code, 5/17. Costs are reported API spend across all 17 tasks. Check out more findings:

DeepSeek V4.1 Flash (1017) solved 8/17 Zero2Repo tasks with dsh, at $3.18 in reported API spend across the full suite. OpenCode scored 7/17 and Claude Code, 5/17. Costs are reported API spend across all 17 tasks. Check out more findings:

30,947 views

Watch the Echo-trained 30B Sokoban model outperform models many times its size, including DeepSeek-R1 and GPT-OSS-120B. Instead of larger datacenters or more expensive GPU fleets, Echo scales reinforcement learning with consumer hardware found everywhere.

Watch the Echo-trained 30B Sokoban model outperform models many times its size, including DeepSeek-R1 and GPT-OSS-120B. Instead of larger datacenters or more expensive GPU fleets, Echo scales reinforcement learning with consumer hardware found everywhere.

75,781 views

Distributed Inference, Now in Hybrid. Try it now:

Distributed Inference, Now in Hybrid. Try it now:

34,803 views

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

A self-evolving agent + a 428B model + 3 Macs = ? Your own AI lab. We ran MiniMax (official) M3 locally with Parallax, right on our desk. Then GenericAgent took over to create a 5-stock portfolio and write it to disk. No cloud. No API bills. Nothing left the machine. Wild to see a ~3K-line agent drive all this with a 400B+ model on local hardware. Thanks to the GenericAgent and MiniMax teams for making local AI feel real.

A self-evolving agent + a 428B model + 3 Macs = ? Your own AI lab. We ran MiniMax (official) M3 locally with Parallax, right on our desk. Then GenericAgent took over to create a 5-stock portfolio and write it to disk. No cloud. No API bills. Nothing left the machine. Wild to see a ~3K-line agent drive all this with a 400B+ model on local hardware. Thanks to the GenericAgent and MiniMax teams for making local AI feel real.

154,509 views • 4 months ago

SWE-bench Verified has become saturated. Seven models scored 95–97% (Vals AI, Sep 2026). But how well can coding agents build complete software projects from scratch? Zero2Repo is a Gradient Data research initiative building hard, verifiable tasks and training data for whole-project development. Agents start with a full spec, an interface contract and an empty workspace. They build the entire repo without reference code, then submit it for hidden deterministic testing. Every test must pass for a task to count. No LLM judge is used. The first release includes 17 tasks derived from real open-source projects across Python, TypeScript, Go, C, C++ and Zig. Across 18 model–harness configurations, scores range from 1/17 to 15/17. The top 10 span 47 percentage points, compared with 5 points on SWE-bench Verified. No configuration has completed all 17 tasks. This matters for self-iteration: small unresolved issues can persist and compound as agents continue developing a project. Reliable self-evolution remains an open research challenge. We’re expanding the task collection and looking for contributors to suggest seed repositories, develop tasks and review requirements and hidden tests. More cases are coming. 👉 👉

SWE-bench Verified has become saturated. Seven models scored 95–97% (Vals AI, Sep 2026). But how well can coding agents build complete software projects from scratch? Zero2Repo is a Gradient Data research initiative building hard, verifiable tasks and training data for whole-project development. Agents start with a full spec, an interface contract and an empty workspace. They build the entire repo without reference code, then submit it for hidden deterministic testing. Every test must pass for a task to count. No LLM judge is used. The first release includes 17 tasks derived from real open-source projects across Python, TypeScript, Go, C, C++ and Zig. Across 18 model–harness configurations, scores range from 1/17 to 15/17. The top 10 span 47 percentage points, compared with 5 points on SWE-bench Verified. No configuration has completed all 17 tasks. This matters for self-iteration: small unresolved issues can persist and compound as agents continue developing a project. Reliable self-evolution remains an open research challenge. We’re expanding the task collection and looking for contributors to suggest seed repositories, develop tasks and review requirements and hidden tests. More cases are coming. 👉 👉

22,843 views • 19 days ago

Reinforcement Learning is the future tense of intelligence. Echo is how it scales. Echo is Gradient’s distributed RL framework, running on everyday consumer devices. From its early experiments, Echo powered a 30B Sokoban model that outperformed DeepSeek-R1 and GPT-OSS-120B.

Reinforcement Learning is the future tense of intelligence. Echo is how it scales. Echo is Gradient’s distributed RL framework, running on everyday consumer devices. From its early experiments, Echo powered a 30B Sokoban model that outperformed DeepSeek-R1 and GPT-OSS-120B.

279,386 views • 1 year ago

The Sovereign AI OS everyone deserves. Parallax lets you set up a local AI cluster across Macs and PCs, and host your own models and applications without performance compromises. Watch Alex Mirran and the crew get Qwen 235B running in minutes.

The Sovereign AI OS everyone deserves. Parallax lets you set up a local AI cluster across Macs and PCs, and host your own models and applications without performance compromises. Watch Alex Mirran and the crew get Qwen 235B running in minutes.

158,182 views • 1 year ago

Introducing Parallax, the first fully distributed inference and serving engine for large language models. Try it now: 🧵

Introducing Parallax, the first fully distributed inference and serving engine for large language models. Try it now: 🧵

161,230 views • 1 year ago

$The future of AI deployment should be fast, affordable, and owned by all. Introducing Gradient Cloud, the all-in-one station for building with AI, powered by the Gradient distributed AI stack. Run frontier models at production speed for a fraction of the cost.$

The future of AI deployment should be fast, affordable, and owned by all. Introducing Gradient Cloud, the all-in-one station for building with AI, powered by the Gradient distributed AI stack. Run frontier models at production speed for a fraction of the cost.

128,424 views • 1 year ago

Builders and believers under one roof at Open Intelligence Day Singapore. We covered deAI infra, GenAI, robotics, trust and ownership, all aligning to power the future of open AGI. Thank you to our partners, speakers, and attendees for making the day a blast. Recap below.

Builders and believers under one roof at Open Intelligence Day Singapore. We covered deAI infra, GenAI, robotics, trust and ownership, all aligning to power the future of open AGI. Thank you to our partners, speakers, and attendees for making the day a blast. Recap below.

71,228 views • 1 year ago

🤯🤯🤯 x402 x Solana is the reel bees knees

🤯🤯🤯 x402 x Solana is the reel bees knees

43,231 views • 1 year ago

No more content to load