Orca 2 · Ollama

orca2

Orca 2 is built by Microsoft research, and are a fine-tuned version of Meta's Llama 2 models. The model is designed to excel particularly in reasoning.

7b 13b

946.3K Pulls 33 Tags Updated 2 years ago

open-orca-platypus2

Merge of the Open Orca OpenChat model and the Garage-bAInd Platypus 2 model. Designed for chat and code generation.

13b

514.6K Pulls 17 Tags Updated 2 years ago

orcarouter/Qwen3.8-27B-Uncensored

Qwen3.8-27B tensor-level abliterated, vision tower and MTP head untouched. 0% over-refusal on XSTest, 0-6% refusal across the A/B suite, no measurable capability loss. Full mmproj vision, tool calling and thinking, 262K context.

vision tools thinking

189.7K Pulls 20 Tags Updated 3 weeks ago

gdisney/orca2-uncensored

Uncensored Orca2 model by Gregory Disney.

2,019 Pulls 1 Tag Updated 2 years ago

Orvyth/engrym-seed

Open-weight local LLM ladder, 2B–27B. Agentic tool calling and thinking. Native 256K context, 32K default. Base 9B scores 91.6% (131/143, current blob). Orvyth seed-tier. Intelligence. Governed.

tools thinking

390 Pulls 18 Tags Updated 1 month ago

meditron

Open-source medical large language model adapted from Llama 2 to the medical domain.

7b 70b

891.8K Pulls 22 Tags Updated 2 years ago

tobestyledintro/oxcoder-9b

OxCoder-9B by OrionLLM: 9B coding model distilled from frontier agent traces (Claude Code / OpenCode / Codex). 262K context, native tool calling, vision. Q4_K_M, Q5_K_M.

vision tools thinking

242 Pulls 2 Tags Updated 1 week ago

SetneufPT/Ornith-1.0-35B-MOE_Q4_110k_24GB-GPU

Custom model for coding with agents to use locally with 24gb or 2x16gb GPUs (working fine...)

tools

210 Pulls 1 Tag Updated 2 months ago

SetneufPT/Ornith-1.0-35B-MOE_Q4_200k_32GB-GPU

Custom model for coding with agents to use locally with 32gb, 2x16gb or 5x8gb GPUs (working very well!!)

tools

98 Pulls 1 Tag Updated 2 months ago

mannix/ornith-1.5-27b-a3b-coder

Code-targeted 184-expert prune of Ornith-1.5-35B-A3B (256e→184e, ~27B, A3B active, top-8). Beats the 256-expert teacher on LiveCodeBench v6 (+6.5pp) and HumanEval+ (+2.4pp) at ~28% fewer experts. Native MTP head + vision tower

vision tools thinking

104 Pulls 39 Tags Updated 1 week ago

satgeze/ornith-35b-1m

Ornith-1.0-35B MoE (APEX-Compact 17GB): 1M context, vision, fits 32GB cards. 50/50 needles to 524K.

vision tools thinking

1,294 Pulls 13 Tags Updated 2 months ago

oamazonasgabriel/qwen3.5-2b

Qwen3.5 2B in Q8_0 quantization. Strong balance of capability and efficiency with 262K context, vision, tool use, and thinking. Ideal for local deployment on consumer hardware.

vision tools thinking

257 Pulls 1 Tag Updated 2 months ago

deepseek-r1

DeepSeek-R1 is a family of open reasoning models with performance approaching that of leading models, such as O3 and Gemini 2.5 Pro.

tools thinking 1.5b 7b 8b 14b 32b 70b 671b

93.1M Pulls 35 Tags Updated 1 year ago

exaone-deep

EXAONE Deep exhibits superior capabilities in various reasoning tasks including math and coding benchmarks, ranging from 2.4B to 32B parameters developed and released by LG AI Research.

2.4b 7.8b 32b

770.3K Pulls 13 Tags Updated 1 year ago

mannix/ornith-1.5-27b-a3b-coderx

REAP-stability-floor 184-expert prune of Ornith-1.5-35B-A3B (~27B, A3B active, top-8) — the recommended arm of the pair. Beats the 256-expert teacher on LiveCodeBench v6 (+10.4pp)

vision tools thinking

278 Pulls 39 Tags Updated 1 week ago

erukude/multiagent-orchestrator

A small multi-agent orchestrator built on Llama3.2 that coordinates LLM agents and tools by outputting "next actions." Use it as the central routing brain in your agentic workflows.

tools 1b 3b

9,243 Pulls 2 Tags Updated 9 months ago

rockypod/public-a11y-coder

Open-source accessibility coding assistant for the public sector. WCAG 2.2 Level AA conformance, Drupal 11, PHP 8.3, Drush 12, Python 3.12, and Playwright (TypeScript) with both axe-core and Siteimprove Alfa.

4b 14b

52 Pulls 2 Tags Updated 4 months ago

PhysicsObsession/blaze-3b

Blaze is a lightweight yet remarkably capable AI model, fine-tuned from the robust foundation of LLaMA 3.2 (3B parameters). Designed with a focus on efficiency and performance. We believe in providing credit to the original creator.

tools

131 Pulls 1 Tag Updated 11 months ago

PhysicsObsession/blaze

Blaze is a lightweight yet remarkably capable AI model, fine-tuned from the robust foundation of LLaMA 3.2 (1B parameters). Designed with a focus on efficiency and performance. We believe in providing credit to the original creator.

tools

50 Pulls 1 Tag Updated 12 months ago

DedeProgames/orion-embedding

Orion Embedding offers a good, comprehensive range of text embedding templates in 2 sizes.

embedding 22m 4b

38 Pulls 2 Tags Updated 6 months ago