@codee_ai

← All tools

Open Models

Flint

Recursive's Apache-2.0 40B mixture-of-experts coding model with 2.5B active parameters.

💡 Built for local agentic coding, at over 125 tokens/s on an M4 Pro MacBook Pro.

Visit Flint

More in Open Models

Open Models

Llama

Meta's family of open-weight models.

💡 Run capable models locally or self-host.
Open Models

DeepSeek

Open-weight reasoning and chat models.

💡 Frontier-level reasoning at very low cost.
Open Models

Qwen

Alibaba's open model family incl. coder models.

💡 Excellent open coding models in many sizes.
Open Models

Hugging Face

The hub for open models, datasets and demos.

💡 Where every open model lands first.
Open Models

Ling-3.1-flash

Ant Group's open 560B hybrid reasoning model.

💡 Open weights for long agent and coding tasks, around 210 tok/s.
Open Models

Liquid AI LFM2.5

24B MoE reasoning model with 2B active parameters.

💡 131k-context local agents that still fit on a laptop.
Open Models

Reflection Beam

Apache-2.0 open-weight model for coding and long-horizon agent work.

💡 Pairs with Mirror CLI, an open-source Rust agent harness.
Open Models

Tess-4.5

Apache-2.0 27B open-weights model for engineering and terminal work.

💡 Runs on a single GPU or a laptop, with a draft head for faster local code generation.
Open Models

Ornith-1.5-122B

MIT-licensed 122B mixture-of-experts model for coding and agent work.

💡 262K context, and the model card reports 84.5 on SWE-bench Verified, on one 80GB GPU.

Get the daily digest

One short email when the digest goes out.