This repository provides a reproduction checkpoint of Axion 1.0-35B, following the recipe described in Axion 1.0: Open Agent Foundation Models, Simplified. The model is designed for scientific reasoning, information seeking, and tool-augmented problem solving.
Alternatives to Axion-1.0-35B
120 open-weight models that do the same job as Axion-1.0-35B (35.1B chat), ranked by how close they are in task and size, then by popularity and how recent they are. Licence and VRAM are shown for each.
Compare Axion-1.0-35B with the top 3Aloha! 🌺 Today, we are releasing Ornith-1.0, a self-improving family of open-source models for agentic coding.
35BA3.8B256K ctx MIT1.5M 518Jun 2026- Same category
- Similar size (35B vs 35.1B)
- Commercial use OK
NVIDIA Nemotron™ is a family of open models with open weights, training data, and recipes, delivering leading efficiency and accuracy for building specialized AI agents.
31.6BA3B256K ctx openmdw-1.1975.1K 274Aug 2026- Same category
- Similar size (31.6B vs 35.1B)
- Commercial use OK
Versioning notice: This card describes the current OTel 2.0 OSFT checkpoint. The model weights may be updated over time. For reproducible evaluation or production deployment, pin a specific model revision, checkpoint hash, or release tag rather than relying on the floating latest version.
31.3B256K ctx Apache 2.0855.4K 32Jul 2026- Same category
- Similar size (31.3B vs 35.1B)
- Commercial use OK
Chirp Chirp! 🐦 We are introducing Ornith-1.5, a major step toward building foundation models through end-to-end self-improvement.
36BA3B256K ctx MIT308.9K 788Aug 2026- Same category
- Similar size (36B vs 35.1B)
- Commercial use OK
Laguna XS 2.1 is a 33B total parameter Mixture-of-Experts model with 3B activated parameters per token designed for agentic coding and long-horizon work on a local machine.
33.4BA2.2B256K ctx Custom licence598.4K 270Jun 2026- Same category
- Similar size (33.4B vs 35.1B)
- Commercial use OK
Laguna XS.2 is a 33B total parameter Mixture-of-Experts model with 3B activated parameters per token designed for agentic coding and long-horizon work on a local machine. It uses Sliding Window Attention with per-head gating in 30 out of 40 layers for fast inference and low KV cache requirements.
33.4BA2.2B256K ctx Apache 2.0743.4K 320Apr 2026- Same category
- Similar size (33.4B vs 35.1B)
- Commercial use OK
Nemotron-3-Nano-30B-A3B-BF16 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response.
31.6BA3B256K ctx nvidia-nemotron-open-model-license1.4M 840Dec 2025- Same category
- Similar size (31.6B vs 35.1B)
- Commercial use OK
👋 Join our Discord community. 📖 Check out the GLM-4.7 technical blog , technical report(GLM-4.5) . 📍 Use GLM-4.7-Flash API services on Z.ai API Platform. 👉 One click to GLM-4.7 .
31.2BA4.6B198K ctx MIT1.1M 1.9KJan 2026- Same category
- Similar size (31.2B vs 35.1B)
- Commercial use OK
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models.
32.8B40K ctx Apache 2.03.2M 794Apr 2025- Same category
- Similar size (32.8B vs 35.1B)
- Commercial use OK
- Widely used
Sage‑A1.5 is a 35B Mixture‑of‑Experts agentic model from Shanghai AI Laboratory, built to advance long-horizon scientific problem solving through adaptive capability composition.
36BA4.7B256K ctx Apache 2.023.9K 405Oct 2026- Same category
- Similar size (36B vs 35.1B)
- Commercial use OK
Flint is Recursive’s mixture-of-experts language model for coding agents, with 39.8 billion total parameters and approximately 2.5 billion active per token. It is designed to investigate codebases, fix bugs, implement features, and write tests.
39.8BA1.3B256K ctx Apache 2.054.7K 1.4KOct 2026- Same category
- Similar size (39.8B vs 35.1B)
- Commercial use OK
- Newer release
gnani-evon-v3.3 is a hybrid Mamba architecture 30B MOE A3B model trained end-to-end for best-in-class Indic capability — continued pretraining, supervised fine-tuning, and reinforcement learning — delivering deep comprehension, extractive QA, and reasoning across Hindi, Bengali, Telugu, Tamil…
31.7BA3B128K ctx Apache 2.068.3K 83Aug 2026- Same category
- Similar size (31.7B vs 35.1B)
- Commercial use OK
In Nanbeige4.3, we take a foundational step toward pushing the boundaries of lightweight agents. Our strategy centers on building a high-capacity teacher agent first, which will subsequently anchor the distillation of advanced agentic capabilities into smaller models.
33.1B512K ctx Apache 2.022.7K 1.1KOct 2026- Same category
- Similar size (33.1B vs 35.1B)
- Commercial use OK
- Newer release
We introduce the updated version of the Qwen3-30B-A3B non-thinking mode, named Qwen3-30B-A3B-Instruct-2507, featuring the following key enhancements:
30.5BA3B256K ctx Apache 2.0820.5K 856Jul 2025- Same category
- Similar size (30.5B vs 35.1B)
- Commercial use OK
K2-Horizon-32B is the large dense member of the K2-Horizon family: a 32B decoder-only model with a 512K context window.
34.8B512K ctx Apache 2.08.7K 42Sept 2026- Same category
- Similar size (34.8B vs 35.1B)
- Commercial use OK
K2-Horizon-MoVA-36B-A4B is the sparse member of the K2-Horizon family: a Mixture-of-Experts model with Mixture-of-Values attention (MoVA) that stores 36B parameters and runs 4B per token. We have released the checkpoints, along with the data and the training code.
37.4BA4B512K ctx Apache 2.015K 435Sept 2026- Same category
- Similar size (37.4B vs 35.1B)
- Commercial use OK
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models.
30.5BA3B40K ctx Apache 2.01.1M 984Apr 2025- Same category
- Similar size (30.5B vs 35.1B)
- Commercial use OK
This is our most spectacular outcome ever. FFT, all parameters, 16bit. 77.4 MMLU on 34b. And it talks like a dream.
34.4B8K ctx Apache 2.04.8M 91May 2024- Same category
- Similar size (34.4B vs 35.1B)
- Commercial use OK
- Widely used
Granite is a family of open-source large language models developed by IBM, designed for enterprise and research use. Granite models are built to be versatile, safe, and efficient — covering a range of sizes and capabilities from compact edge-deployable models to large-scale reasoning systems.
29.3B128K ctx Apache 2.037.2K 143Aug 2026- Same category
- Similar size (29.3B vs 35.1B)
- Commercial use OK
Qwen-AgentWorld is the first language world model to cover seven agent interaction domains within a single model. It simulates agentic environments via long chain-of-thought reasoning, predicting the next environment state given an agent's action and interaction history.
34.7BA3B256K ctx Apache 2.09.6K 749Jun 2026- Same category
- Similar size (34.7B vs 35.1B)
- Commercial use OK
MLExpert is a 35B Mixture-of-Experts agentic model for autonomous machine learning engineering. It organizes experiments, writes and revises code, and improves solutions through sustained interaction with execution environments.
34.7BA3.5B256K ctx Apache 2.03.7K 17Oct 2026- Same category
- Similar size (34.7B vs 35.1B)
- Commercial use OK
Agents‑A1 is a 35B Mixture‑of‑Experts agentic model from InternScience, built to scale heterogeneous agentic abilities across multiple domains including Long‑horizon Search, Engineering, Scientific Research, Instruction Following, and Tool-calling.
35.1BA3.9B256K ctx Apache 2.05.7K 669Jun 2026- Same category
- Similar size (35.1B vs 35.1B)
- Commercial use OK
A next-generation family of agentic models built for long-horizon tasks in real-world environments.
35.1BA3.9B256K ctx Apache 2.02.9K 883Sept 2026- Same category
- Similar size (35.1B vs 35.1B)
- Commercial use OK
For full project documentation, deployment instructions, SDK usage, and data details, visit the GitHub repository.
31.2BA5.5B256K ctx Apache 2.06.1K 327Oct 2026- Same category
- Similar size (31.2B vs 35.1B)
- Commercial use OK