DeepSeek-V4.1-Flash is live on Featherless

DeepSeek's 763B multimodal MoE tops Opus 5.0 on short-horizon agent benchmarks, and it is live on Featherless at 256K context in FP8.

Kimi K3 is live on Featherless

Moonshot's Kimi K3 is the first open-weight model at nearly 3 trillion parameters. What it scores, what it takes to serve, and how to run it on Featherless.

GLM 5.2 vs Opus for Agent Workloads (Reduce cost from $150K to $7.6K with AMD GPUs)

Dedicated GLM 5.2 on AMD MI325X cuts a $150K/month Claude Opus agent bill to $7.6K, and the Fable shutdown shows why open models matter for continuity.

Key Value Means Transformers with Expandable block compressed memory

KVM turns the KV cache into a compressed, expandable memory. It bridges linear RNNs and full attention, keeping inference efficient without losing recall.

Featherless Raises $20M Series A to Power Open-Source AI Infrastructure

Featherless has raised a $20 million Series A round to accelerate one mission: making open-source AI practical, sovereign, reliable and effortless at scale.

Keep Creating with Featherless

Introducing the Builder Series: a long-running spotlight on people doing meaningful work with open-source AI — with free Featherless access, technical and awareness support, and community.

No Language Forgotten

Backing the people building AI for every language‍

Open-Source AI Agents Now Have a Home

Today we're launching a managed runtime for open-source AI agents, starting with OpenClaw.

MiniMax M2.5 Now Available on Featherless!

MiniMax M2.5 is now live on Featherless. It's a frontier-class reasoning model that tops SWE-Bench Verified at 80.2%, available now through our API.

Using Featherless.ai Models with OpenClaw

A step-by-step guide to using Featherless.ai models in OpenClaw via OpenAI-compatible configuration

DeepSeek V3.2 & V3.2 Speciale Now Available on Featherless!

DeepSeek V3.2 matches GPT-5 and V3.2 Speciale passes it, taking gold at IMO and IOI 2025. Both are now available on Featherless with unlimited access.

Featherless.ai & Z.ai launch GLM 5.2 access worldwide