Curated picks of the strongest open-source AI models, ranked by use case.
The strongest open-weight large language models of 2026, ranked by capability, deployment economics, licensing, and real-world reliability — based on the current state of the leaderboards in April 2026.
The strongest open-weight models for coding workloads in 2026 — agentic coding, code completion, code review, and full-codebase reasoning — ranked by SWE-Bench performance, deployment economics, and real-world reliability.
The strongest open-weight models for extended chain-of-thought reasoning, mathematical problem solving, and structured analysis — ranked across AIME, GPQA, and complex code generation benchmarks.
The strongest small open-weight models for on-device, edge, and consumer-hardware deployment in 2026 — ranked by quality at 4B, 7B, and 14B parameter scales for local inference on phones, laptops, and desktop GPUs.
The strongest open-weight base models for QLoRA and LoRA fine-tuning in 2026 — ranked by hardware accessibility, quality of the resulting fine-tunes, ecosystem support, and licensing for commercial deployment.
The strongest open-weight models for agentic workloads in 2026 — multi-step planning, tool use, function calling, and long-horizon execution — ranked by reliability in real agentic deployments rather than synthetic benchmarks.
The strongest open-weight models that natively accept image, audio, or video input alongside text — ranked by capability, deployment economics, and licensing for production multimodal applications.
The strongest open-weight models with minimal refusal training — well-suited to legitimate use cases like security research, red-team evaluation, mature creative writing, and educational discussion of sensitive topics where mainstream models' over-refusal is an obstacle.
The strongest open-weight models for running locally on Apple Silicon Macs (M1/M2/M3/M4) — ranked by quality, MLX support, and memory footprint for typical Mac configurations from 16GB MacBook Air to 192GB Mac Studio.
The strongest open-weight models that fit in under 10GB of VRAM at standard Q4_K_M quantization — for laptop GPUs, RTX 3060/4060 12GB cards, and any deployment where memory is the binding constraint.
The strongest open-weight models for retrieval-augmented generation in 2026 — ranked by long-context retrieval quality, instruction-following stability, and inference economics for production RAG pipelines.
The strongest open-weight models with 1M+ token context windows in 2026 — ranked by effective context retention, architecture efficiency, and practical deployment for full-codebase or long-document reasoning.