Articles tagged "MIT License"

Ornith 1.5 is an MIT open-weight LLM needing 8GB to 800GB VRAM. Compare GPU picks and quantized VRAM needs for 9B, 35B-A3B and 397B. As of August 2026.

Prime Agent treats context as a variable and sub-agents as function calls via its Recursive Language Model design, paired with a Continual Harness you can CRUD-edit. Here's how Prime Intellect's MIT-licensed coding agent actually works. Updated Aug 2026.

OpenChamber is an open-source agentic dev environment on OpenCode, hitting HN in Aug 2026. This guide covers Session Goals, multi-model runs, and setup steps.

LongCat-2.0 is Meituan's 1.6T-parameter, MIT-licensed MoE model. Running it locally needs roughly 3,800GB in BF16 or about 970GB even at INT4 — no personal PC can hold it. Requirements tables and API pricing ($0.75/$2.95 per 1M tokens), updated July 2026.

DeepSeek V4-Flash needs roughly 160GB at 4-bit and V4-Pro about 920GB. The July 31 V4-Flash-0731 build keeps the same footprint while beating V4-Pro on agent benchmarks. VRAM, RAM and GPU tables by quantization, API pricing, and the legacy model retirement. Updated August 2026.

GLM-5.2 needs roughly 430GB at 4-bit and about 1.5TB at BF16 in combined memory. Quick-lookup VRAM, RAM and quantization tables for running Z.ai's 753B / ~40B-active open-weight MoE (MIT-licensed) locally. Updated July 2026.

**[Hunk](https://www.hunk.dev/) is a terminal-based diff viewer purpose-built for code review and AI-agent integration** — a replacement for stock `git diff`, with review-experience as a first-class design goal. **Key features**: (1) **multi-file review sidebar** — see the whole changeset, navigate files, view change counts, (2) **inline AI annotations** — reasoning from AI agents ([Claude Code](https://claude.com/product/claude-code), [Cursor](../columns/cursor-ios-supports-2026-07-01), etc.) shows adjacent to the relevant code, (3) **adaptive layout** — split / stacked / auto-responsive views auto-adjust to terminal width, (4) **watch mode** — auto-refreshes as the working tree changes, (5) **syntax highlighting + themes** — Graphite / Midnight / Ember / Zenburn / Catppuccin, (6) multiple input methods — **keyboard, mouse, pager integration**. **CLI commands**: `hunk diff` (uncommitted changes) and `hunk show` (a commit). **Tech stack**: Node.js 18+, **MIT license**, distributed via **npm / Homebrew / Nix**. Position: the terminal counterpart to [Crit.md's browser PR-review experience](../columns/crit-md-local-first-agent-review-2026-07); alongside [Herdr (parallel agent operations)](../columns/herdr-terminal-agent-multiplexer-2026-07), it's part of the same "terminal-native AI-agent era" trend. **Target users**: developers doing code reviews, teams integrating AI tools into dev workflows, and terminal-first engineers.

**DeepReinforce released Ornith-1.0 on June 26, 2026** ([official](https://deep-reinforce.com/ornith_1_0.html) / [Hugging Face collection](https://huggingface.co/collections/deepreinforce-ai/ornith-10)). It is an **MIT-licensed open-weights family specialized for agentic coding**, **with no regional restrictions**. **Three sizes**: [Ornith-1.0-9B](https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B) (dense, ~19GB BF16) / [Ornith-1.0-35B](https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B) (MoE) / [Ornith-1.0-397B](https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B) (MoE, built on Qwen 3.5 + Gemma 4). **All sizes ship 262K context**, with **FP8 and GGUF quantizations released alongside**. **Benchmarks (vendor-reported, claimed SOTA at each open-weights size tier)**: | Benchmark | 9B | 35B | 397B | |---|---|---|---| | **SWE-Bench Verified** | **69.4%** | **75.6%** | **82.4%** | | **SWE-Bench Pro** | **42.9%** | **50.4%** | **62.2%** | | **SWE-Bench Multilingual** | — | — | **78.9%** | | **Terminal-Bench 2.1** | 43.1% | 64.2% | **77.5-78.2%** | | **NL2Repo** | 27.2% | 34.6% | **48.2%** | | **ClawEval** | — | — | **77.1%** | **Design thesis**: Reinforcement learning optimizes **both the solution rollouts and the scaffolding (the agent structure that drives them) itself** — a 'self-improving' agentic-coding design. It sits naturally next to the [Loop Engineering Maker-Checker](../columns/loop-engineering-ai-agent-paradigm-2026-06) paradigm. Reasoning is exposed via `<think>...</think>` blocks; function calling and tool use are first-class. **Distribution and ops**: vLLM ≥ 0.19.1 / SGLang ≥ 0.5.9 / Transformers ≥ 5.8.1 / Docker + llama.cpp / Ollama. OpenAI-compatible endpoints. The 9B fits on a single 80GB GPU; 35B and 397B want an **8×80GB GPU node (TP=8)**. Agent-framework compatibility: **OpenHands, Hermes Agent, and [OpenClaw](../services/openclaw-setup)** (Oflight's own service line — and ClawEval is in DeepReinforce's published benchmark set). **DeepReinforce lineage**: an RL-focused research organization that has previously shipped [CUDA-L1 (avg 3.12× GPU speedup)](https://github.com/deepreinforce-ai/CUDA-L1), [CUDA-L2 (HGEMM kernels beating cuBLAS)](https://github.com/deepreinforce-ai/CUDA-L2), and **IterX (MLSys 2026 NVIDIA Track)**. Ornith-1.0 applies the same RL playbook to LLM self-improvement. **Positioning**: alongside [Kimi K2.7-Code](../columns/kimi-k2-7-code-moonshot-ai-2026-06) (1T MoE / 32B active) and [GLM-5.2](../columns/local-llm-landscape-2026-june-update) (Intelligence Index v4.1 = 51, open-weights leader), **Ornith-1.0 is at the front of the June-2026 agentic-coding open-weights race**. Against Chinese-origin models (Kimi / GLM), its differentiator is **MIT license + no regional restrictions + a US-flag procurement story**. **Caveat**: benchmarks are DeepReinforce's own vendor-reported numbers. Independent third-party verification on public leaderboards has not yet appeared (as of June 26, 2026). The article closes with **three inquiry funnels for Ornith-1.0–era local-LLM evaluation, build, and ongoing maintenance**.

Kimi K2.5, released by Moonshot AI on January 27, 2026, is a 1 trillion parameter (32B active) MoE model under the MIT License. It scores 76.8% on SWE-bench, 99.0% on HumanEval, and 87.6% on GPQA Diamond. This guide covers its architecture, hardware requirements, Ollama setup, and practical use cases.