Alternatives to Tülu 3
Compare Tülu 3 alternatives for your business or organization using the curated list below. SourceForge ranks the best alternatives to Tülu 3 in 2026. Compare features, ratings, user reviews, pricing, and more from Tülu 3 competitors and alternatives in order to make an informed decision for your business.
-
1
Olmo 3
Ai2
Olmo 3 is a fully open model family spanning 7 billion and 32 billion parameter variants that delivers not only high-performing base, reasoning, instruction, and reinforcement-learning models, but also exposure of the entire model flow, including raw training data, intermediate checkpoints, training code, long-context support (65,536 token window), and provenance tooling. Starting with the Dolma 3 dataset (≈9 trillion tokens) and its disciplined mix of web text, scientific PDFs, code, and long-form documents, the pre-training, mid-training, and long-context phases shape the base models, which are then post-trained via supervised fine-tuning, direct preference optimisation, and RL with verifiable rewards to yield the Think and Instruct variants. The 32 B Think model is described as the strongest fully open reasoning model to date, competitively close to closed-weight peers in math, code, and complex reasoning.Starting Price: Free -
2
Molmo
Ai2
Molmo is a family of open, state-of-the-art multimodal AI models developed by the Allen Institute for AI (Ai2). These models are designed to bridge the gap between open and proprietary systems, achieving competitive performance across a wide range of academic benchmarks and human evaluations. Unlike many existing multimodal models that rely heavily on synthetic data from proprietary systems, Molmo is trained entirely on open data, ensuring transparency and reproducibility. A key innovation in Molmo's development is the introduction of PixMo, a novel dataset comprising highly detailed image captions collected from human annotators using speech-based descriptions, as well as 2D pointing data that enables the models to answer questions using both natural language and non-verbal cues. This allows Molmo to interact with its environment in more nuanced ways, such as pointing to objects within images, thereby enhancing its applicability in fields like robotics and augmented reality. -
3
Llama 2
Meta
The next generation of our open source large language model. This release includes model weights and starting code for pretrained and fine-tuned Llama language models — ranging from 7B to 70B parameters. Llama 2 pretrained models are trained on 2 trillion tokens, and have double the context length than Llama 1. Its fine-tuned models have been trained on over 1 million human annotations. Llama 2 outperforms other open source language models on many external benchmarks, including reasoning, coding, proficiency, and knowledge tests. Llama 2 was pretrained on publicly available online data sources. The fine-tuned model, Llama-2-chat, leverages publicly available instruction datasets and over 1 million human annotations. We have a broad range of supporters around the world who believe in our open approach to today’s AI — companies that have given early feedback and are excited to build with Llama 2.Starting Price: Free -
4
Code Llama
Meta
Code Llama is a large language model (LLM) that can use text prompts to generate code. Code Llama is state-of-the-art for publicly available LLMs on code tasks, and has the potential to make workflows faster and more efficient for current developers and lower the barrier to entry for people who are learning to code. Code Llama has the potential to be used as a productivity and educational tool to help programmers write more robust, well-documented software. Code Llama is a state-of-the-art LLM capable of generating code, and natural language about code, from both code and natural language prompts. Code Llama is free for research and commercial use. Code Llama is built on top of Llama 2 and is available in three models: Code Llama, the foundational code model; Codel Llama - Python specialized for Python; and Code Llama - Instruct, which is fine-tuned for understanding natural language instructions.Starting Price: Free -
5
Olmo 2
Ai2
Olmo 2 is a family of fully open language models developed by the Allen Institute for AI (AI2), designed to provide researchers and developers with transparent access to training data, open-source code, reproducible training recipes, and comprehensive evaluations. These models are trained on up to 5 trillion tokens and are competitive with leading open-weight models like Llama 3.1 on English academic benchmarks. Olmo 2 emphasizes training stability, implementing techniques to prevent loss spikes during long training runs, and utilizes staged training interventions during late pretraining to address capability deficiencies. The models incorporate state-of-the-art post-training methodologies from AI2's Tülu 3, resulting in the creation of Olmo 2-Instruct models. An actionable evaluation framework, the Open Language Modeling Evaluation System (OLMES), was established to guide improvements through development stages, consisting of 20 evaluation benchmarks assessing core capabilities. -
6
LongLLaMA
LongLLaMA
This repository contains the research preview of LongLLaMA, a large language model capable of handling long contexts of 256k tokens or even more. LongLLaMA is built upon the foundation of OpenLLaMA and fine-tuned using the Focused Transformer (FoT) method. LongLLaMA code is built upon the foundation of Code Llama. We release a smaller 3B base variant (not instruction tuned) of the LongLLaMA model on a permissive license (Apache 2.0) and inference code supporting longer contexts on hugging face. Our model weights can serve as the drop-in replacement of LLaMA in existing implementations (for short context up to 2048 tokens). Additionally, we provide evaluation results and comparisons against the original OpenLLaMA models.Starting Price: Free -
7
OpenPipe
OpenPipe
OpenPipe provides fine-tuning for developers. Keep your datasets, models, and evaluations all in one place. Train new models with the click of a button. Automatically record LLM requests and responses. Create datasets from your captured data. Train multiple base models on the same dataset. We serve your model on our managed endpoints that scale to millions of requests. Write evaluations and compare model outputs side by side. Change a couple of lines of code, and you're good to go. Simply replace your Python or Javascript OpenAI SDK and add an OpenPipe API key. Make your data searchable with custom tags. Small specialized models cost much less to run than large multipurpose LLMs. Replace prompts with models in minutes, not weeks. Fine-tuned Mistral and Llama 2 models consistently outperform GPT-4-1106-Turbo, at a fraction of the cost. We're open-source, and so are many of the base models we use. Own your own weights when you fine-tune Mistral and Llama 2, and download them at any time.Starting Price: $1.20 per 1M tokens -
8
Qwen2
Alibaba
Qwen2 is the large language model series developed by Qwen team, Alibaba Cloud. Qwen2 is a series of large language models developed by the Qwen team at Alibaba Cloud. It includes both base language models and instruction-tuned models, ranging from 0.5 billion to 72 billion parameters, and features both dense models and a Mixture-of-Experts model. The Qwen2 series is designed to surpass most previous open-weight models, including its predecessor Qwen1.5, and to compete with proprietary models across a broad spectrum of benchmarks in language understanding, generation, multilingual capabilities, coding, mathematics, and reasoning.Starting Price: Free -
9
Tinker
Thinking Machines Lab
Tinker is a training API designed for researchers and developers that allows full control over model fine-tuning while abstracting away the infrastructure complexity. It supports primitives and enables users to build custom training loops, supervision logic, and reinforcement learning flows. It currently supports LoRA fine-tuning on open-weight models across both LLama and Qwen families, ranging from small models to large mixture-of-experts architectures. Users write Python code to handle data, loss functions, and algorithmic logic; Tinker handles scheduling, resource allocation, distributed training, and failure recovery behind the scenes. The service lets users download model weights at different checkpoints and doesn’t force them to manage the compute environment. Tinker is delivered as a managed offering; training jobs run on Thinking Machines’ internal GPU infrastructure, freeing users from cluster orchestration. -
10
Mistral 7B
Mistral AI
Mistral 7B is a 7.3-billion-parameter language model that outperforms larger models like Llama 2 13B across various benchmarks. It employs Grouped-Query Attention (GQA) for faster inference and Sliding Window Attention (SWA) to efficiently handle longer sequences. Released under the Apache 2.0 license, Mistral 7B is accessible for deployment across diverse platforms, including local environments and major cloud services. Additionally, a fine-tuned version, Mistral 7B Instruct, demonstrates enhanced performance in instruction-following tasks, surpassing models like Llama 2 13B Chat.Starting Price: Free -
11
Qwen2.5-Max
Alibaba
Qwen2.5-Max is a large-scale Mixture-of-Experts (MoE) model developed by the Qwen team, pretrained on over 20 trillion tokens and further refined through Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF). In evaluations, it outperforms models like DeepSeek V3 in benchmarks such as Arena-Hard, LiveBench, LiveCodeBench, and GPQA-Diamond, while also demonstrating competitive results in other assessments, including MMLU-Pro. Qwen2.5-Max is accessible via API through Alibaba Cloud and can be explored interactively on Qwen Chat.Starting Price: Free -
12
Phi-4-reasoning
Microsoft
Phi-4-reasoning is a 14-billion parameter transformer-based language model optimized for complex reasoning tasks, including math, coding, algorithmic problem solving, and planning. Trained via supervised fine-tuning of Phi-4 on carefully curated "teachable" prompts and reasoning demonstrations generated using o3-mini, it generates detailed reasoning chains that effectively leverage inference-time compute. Phi-4-reasoning incorporates outcome-based reinforcement learning to produce longer reasoning traces. It outperforms significantly larger open-weight models such as DeepSeek-R1-Distill-Llama-70B and approaches the performance levels of the full DeepSeek-R1 model across a wide range of reasoning tasks. Phi-4-reasoning is designed for environments with constrained computing or latency. Fine-tuned with synthetic data generated by DeepSeek-R1, it provides high-quality, step-by-step problem solving. -
13
Alpaca
Stanford Center for Research on Foundation Models (CRFM)
Instruction-following models such as GPT-3.5 (text-DaVinci-003), ChatGPT, Claude, and Bing Chat have become increasingly powerful. Many users now interact with these models regularly and even use them for work. However, despite their widespread deployment, instruction-following models still have many deficiencies: they can generate false information, propagate social stereotypes, and produce toxic language. To make maximum progress on addressing these pressing problems, it is important for the academic community to engage. Unfortunately, doing research on instruction-following models in academia has been difficult, as there is no easily accessible model that comes close in capabilities to closed-source models such as OpenAI’s text-DaVinci-003. We are releasing our findings about an instruction-following language model, dubbed Alpaca, which is fine-tuned from Meta’s LLaMA 7B model. -
14
StableVicuna
Stability AI
StableVicuna is the first large-scale open source chatbot trained via reinforced learning from human feedback (RHLF). StableVicuna is a further instruction fine tuned and RLHF trained version of Vicuna v0 13b, which is an instruction fine tuned LLaMA 13b model. In order to achieve StableVicuna’s strong performance, we utilize Vicuna as the base model and follow the typical three-stage RLHF pipeline outlined by Steinnon et al. and Ouyang et al. Concretely, we further train the base Vicuna model with supervised finetuning (SFT) using a mixture of three datasets: OpenAssistant Conversations Dataset (OASST1), a human-generated, human-annotated assistant-style conversation corpus comprising 161,443 messages distributed across 66,497 conversation trees, in 35 different languages; GPT4All Prompt Generations, a dataset of 437,605 prompts and responses generated by GPT-3.5 Turbo; And Alpaca, a dataset of 52,000 instructions and demonstrations generated by OpenAI's text-davinci-003.Starting Price: Free -
15
Sky-T1
NovaSky
Sky-T1-32B-Preview is an open source reasoning model developed by the NovaSky team at UC Berkeley's Sky Computing Lab. It matches the performance of proprietary models like o1-preview on reasoning and coding benchmarks, yet was trained for under $450, showcasing the feasibility of cost-effective, high-level reasoning capabilities. The model was fine-tuned from Qwen2.5-32B-Instruct using a curated dataset of 17,000 examples across diverse domains, including math and coding. The training was completed in 19 hours on eight H100 GPUs with DeepSpeed Zero-3 offloading. All aspects of the project, including data, code, and model weights, are fully open-source, empowering the academic and open-source communities to replicate and enhance the model's performance.Starting Price: Free -
16
StarCoder
BigCode
StarCoder and StarCoderBase are Large Language Models for Code (Code LLMs) trained on permissively licensed data from GitHub, including from 80+ programming languages, Git commits, GitHub issues, and Jupyter notebooks. Similar to LLaMA, we trained a ~15B parameter model for 1 trillion tokens. We fine-tuned StarCoderBase model for 35B Python tokens, resulting in a new model that we call StarCoder. We found that StarCoderBase outperforms existing open Code LLMs on popular programming benchmarks and matches or surpasses closed models such as code-cushman-001 from OpenAI (the original Codex model that powered early versions of GitHub Copilot). With a context length of over 8,000 tokens, the StarCoder models can process more input than any other open LLM, enabling a wide range of interesting applications. For example, by prompting the StarCoder models with a series of dialogues, we enabled them to act as a technical assistant.Starting Price: Free -
17
NVIDIA Llama Nemotron
NVIDIA
NVIDIA Llama Nemotron is a family of advanced language models optimized for reasoning and a diverse set of agentic AI tasks. These models excel in graduate-level scientific reasoning, advanced mathematics, coding, instruction following, and tool calls. Designed for deployment across various platforms, from data centers to PCs, they offer the flexibility to toggle reasoning capabilities on or off, reducing inference costs when deep reasoning isn't required. The Llama Nemotron family includes models tailored for different deployment needs. Built upon Llama models and enhanced by NVIDIA through post-training, these models demonstrate improved accuracy, up to 20% over base models, and optimized inference speeds, achieving up to five times the performance of other leading open reasoning models. This efficiency enables handling more complex reasoning tasks, enhances decision-making capabilities, and reduces operational costs for enterprises. -
18
EXAONE Deep
LG
EXAONE Deep is a series of reasoning-enhanced language models developed by LG AI Research, featuring parameter sizes of 2.4 billion, 7.8 billion, and 32 billion. These models demonstrate superior capabilities in various reasoning tasks, including math and coding benchmarks. Notably, EXAONE Deep 2.4B outperforms other models of comparable size, EXAONE Deep 7.8B surpasses both open-weight models of similar scale and the proprietary reasoning model OpenAI o1-mini, and EXAONE Deep 32B shows competitive performance against leading open-weight models. The repository provides comprehensive documentation covering performance evaluations, quickstart guides for using EXAONE Deep models with Transformers, explanations of quantized EXAONE Deep weights in AWQ and GGUF formats, and instructions for running EXAONE Deep models locally using frameworks like llama.cpp and Ollama.Starting Price: Free -
19
Vicuna
lmsys.org
Vicuna-13B is an open-source chatbot trained by fine-tuning LLaMA on user-shared conversations collected from ShareGPT. Preliminary evaluation using GPT-4 as a judge shows Vicuna-13B achieves more than 90%* quality of OpenAI ChatGPT and Google Bard while outperforming other models like LLaMA and Stanford Alpaca in more than 90%* of cases. The cost of training Vicuna-13B is around $300. The code and weights, along with an online demo, are publicly available for non-commercial use.Starting Price: Free -
20
Kimi K2
Moonshot AI
Kimi K2 is a state-of-the-art open source large language model series built on a mixture-of-experts (MoE) architecture, featuring 1 trillion total parameters and 32 billion activated parameters for task-specific efficiency. Trained with the Muon optimizer on over 15.5 trillion tokens and stabilized by MuonClip’s attention-logit clamping, it delivers exceptional performance in frontier knowledge, reasoning, mathematics, coding, and general agentic workflows. Moonshot AI provides two variants, Kimi-K2-Base for research-level fine-tuning and Kimi-K2-Instruct pre-trained for immediate chat and tool-driven interactions, enabling both custom development and drop-in agentic capabilities. Benchmarks show it outperforms leading open source peers and rivals top proprietary models in coding tasks and complex task breakdowns, while its 128 K-token context length, tool-calling API compatibility, and support for industry-standard inference engines.Starting Price: Free -
21
Hermes 3
Nous Research
Experiment, and push the boundaries of individual alignment, artificial consciousness, open-source software, and decentralization, in ways that monolithic companies and governments are too afraid to try. Hermes 3 contains advanced long-term context retention and multi-turn conversation capability, complex roleplaying and internal monologue abilities, and enhanced agentic function-calling. Our training data aggressively encourages the model to follow the system and instruction prompts exactly and in an adaptive manner. Hermes 3 was created by fine-tuning Llama 3.1 8B, 70B, and 405B, and training on a dataset of primarily synthetically generated responses. The model boasts comparable and superior performance to Llama 3.1 while unlocking deeper capabilities in reasoning and creativity. Hermes 3 is a series of instruct and tool-use models with strong reasoning and creative abilities.Starting Price: Free -
22
LFM2.5
Liquid AI
Liquid AI’s LFM2.5 is the next generation of on-device AI foundation models designed to deliver high-performance, efficient AI inference on edge devices such as phones, laptops, vehicles, IoT systems, and embedded hardware without relying on cloud compute. It extends the previous LFM2 architecture by significantly increasing the pretraining scale and reinforcement learning stages, yielding a family of hybrid models around 1.2 billion parameters that balance instruction following, reasoning, and multimodal capabilities for real-world agentic use cases. The LFM2.5 family includes Base (for fine-tuning and customization), Instruct (general-purpose instruction-tuned), Japanese-optimized, Vision-Language, and Audio-Language variants, all optimized for fast, on-device inference under tight memory constraints and available as open-weight models deployable via frameworks like llama.cpp, MLX, vLLM, and ONNX.Starting Price: Free -
23
SERA
Ai2
Open Coding Agents are a family of fully open, high-performance AI coding models and an associated training method released by the Allen Institute for AI that make building, customizing, and training coding agents on any repository remarkably accessible, affordable, and transparent; the platform includes models, code, training recipes, and tools that can be launched with minimal setup so users can tailor agents to their own codebases and engineering conventions for tasks like code generation, code review, debugging, maintenance, and code explanation. These agents break from the traditional closed, expensive systems by offering an open pipeline from models to training data and enabling fine-tuning on internal code to teach agents about organization-specific APIs, patterns, and workflows; the first release, SERA (Soft-verified Efficient Repository Agents), achieves state-of-the-art performance on coding benchmarks at a fraction of the typical compute cost.Starting Price: Free -
24
Oumi
Oumi
Oumi is a fully open source platform that streamlines the entire lifecycle of foundation models, from data preparation and training to evaluation and deployment. It supports training and fine-tuning models ranging from 10 million to 405 billion parameters using state-of-the-art techniques such as SFT, LoRA, QLoRA, and DPO. The platform accommodates both text and multimodal models, including architectures like Llama, DeepSeek, Qwen, and Phi. Oumi offers tools for data synthesis and curation, enabling users to generate and manage training datasets effectively. For deployment, it integrates with popular inference engines like vLLM and SGLang, ensuring efficient model serving. The platform also provides comprehensive evaluation capabilities across standard benchmarks to assess model performance. Designed for flexibility, Oumi can run on various environments, from local laptops to cloud infrastructures such as AWS, Azure, GCP, and Lambda.Starting Price: Free -
25
Mistral Large 3
Mistral AI
Mistral Large 3 is a next-generation, open multimodal AI model built with a powerful sparse Mixture-of-Experts architecture featuring 41B active parameters out of 675B total. Designed from scratch on NVIDIA H200 GPUs, it delivers frontier-level reasoning, multilingual performance, and advanced image understanding while remaining fully open-weight under the Apache 2.0 license. The model achieves top-tier results on modern instruction benchmarks, positioning it among the strongest permissively licensed foundation models available today. With native support across vLLM, TensorRT-LLM, and major cloud providers, Mistral Large 3 offers exceptional accessibility and performance efficiency. Its design enables enterprise-grade customization, letting teams fine-tune or adapt the model for domain-specific workflows and proprietary applications. Mistral Large 3 represents a major advancement in open AI, offering frontier intelligence without sacrificing transparency or control.Starting Price: Free -
26
Stable Beluga
Stability AI
Stability AI and its CarperAI lab proudly announce Stable Beluga 1 and its successor Stable Beluga 2 (formerly codenamed FreeWilly), two powerful new, open access, Large Language Models (LLMs). Both models demonstrate exceptional reasoning ability across varied benchmarks. Stable Beluga 1 leverages the original LLaMA 65B foundation model and was carefully fine-tuned with a new synthetically-generated dataset using Supervised Fine-Tune (SFT) in standard Alpaca format. Similarly, Stable Beluga 2 leverages the LLaMA 2 70B foundation model to achieve industry-leading performance.Starting Price: Free -
27
Phi-4-mini-reasoning
Microsoft
Phi-4-mini-reasoning is a 3.8-billion parameter transformer-based language model optimized for mathematical reasoning and step-by-step problem solving in environments with constrained computing or latency. Fine-tuned with synthetic data generated by the DeepSeek-R1 model, it balances efficiency with advanced reasoning ability. Trained on over one million diverse math problems spanning multiple levels of difficulty from middle school to Ph.D. level, Phi-4-mini-reasoning outperforms its base model on long sentence generation across various evaluations and surpasses larger models like OpenThinker-7B, Llama-3.2-3B-instruct, and DeepSeek-R1. It features a 128K-token context window and supports function calling, enabling integration with external tools and APIs. Phi-4-mini-reasoning can be quantized using Microsoft Olive or Apple MLX Framework for deployment on edge devices such as IoT, laptops, and mobile devices. -
28
kluster.ai
kluster.ai
Kluster.ai is a developer-centric AI cloud platform designed to deploy, scale, and fine-tune large language models (LLMs) with speed and efficiency. Built for developers by developers, it offers Adaptive Inference, a flexible and scalable service that adjusts seamlessly to workload demands, ensuring high-performance processing and consistent turnaround times. Adaptive Inference provides three distinct processing options: real-time inference for ultra-low latency needs, asynchronous inference for cost-effective handling of flexible timing tasks, and batch inference for efficient processing of high-volume, bulk tasks. It supports a range of open-weight, cutting-edge multimodal models for chat, vision, code, and more, including Meta's Llama 4 Maverick and Scout, Qwen3-235B-A22B, DeepSeek-R1, and Gemma 3 . Kluster.ai's OpenAI-compatible API allows developers to integrate these models into their applications seamlessly.Starting Price: $0.15per input -
29
MPT-7B
MosaicML
Introducing MPT-7B, the latest entry in our MosaicML Foundation Series. MPT-7B is a transformer trained from scratch on 1T tokens of text and code. It is open source, available for commercial use, and matches the quality of LLaMA-7B. MPT-7B was trained on the MosaicML platform in 9.5 days with zero human intervention at a cost of ~$200k. Now you can train, finetune, and deploy your own private MPT models, either starting from one of our checkpoints or training from scratch. For inspiration, we are also releasing three finetuned models in addition to the base MPT-7B: MPT-7B-Instruct, MPT-7B-Chat, and MPT-7B-StoryWriter-65k+, the last of which uses a context length of 65k tokens!Starting Price: Free -
30
Nebius Token Factory
Nebius
Nebius Token Factory is a scalable AI inference platform designed to run open-source and custom AI models in production without manual infrastructure management. It offers enterprise-ready inference endpoints with predictable performance, autoscaling throughput, and sub-second latency — even at very high request volumes. It delivers 99.9% uptime availability and supports unlimited or tailored traffic profiles based on workload needs, simplifying the transition from experimentation to global deployment. Nebius Token Factory supports a broad set of open source models such as Llama, Qwen, DeepSeek, GPT-OSS, Flux, and many others, and lets teams host and fine-tune models through an API or dashboard. Users can upload LoRA adapters or full fine-tuned variants directly, with the same enterprise performance guarantees applied to custom models.Starting Price: $0.02 -
31
Llama 3.1
Meta
The open source AI model you can fine-tune, distill and deploy anywhere. Our latest instruction-tuned model is available in 8B, 70B and 405B versions. Using our open ecosystem, build faster with a selection of differentiated product offerings to support your use cases. Choose from real-time inference or batch inference services. Download model weights to further optimize cost per token. Adapt for your application, improve with synthetic data and deploy on-prem or in the cloud. Use Llama system components and extend the model using zero shot tool use and RAG to build agentic behaviors. Leverage 405B high quality data to improve specialized models for specific use cases.Starting Price: Free -
32
DeepCoder
Agentica Project
DeepCoder is a fully open source code-reasoning and generation model released by Agentica Project in collaboration with Together AI. It is fine-tuned from DeepSeek-R1-Distilled-Qwen-14B using distributed reinforcement learning, achieving a 60.6% accuracy on LiveCodeBench (representing an 8% improvement over the base), a performance level that matches that of proprietary models such as o3-mini (2025-01-031 Low) and o1 while using only 14 billion parameters. It was trained over 2.5 weeks on 32 H100 GPUs with a curated dataset of roughly 24,000 coding problems drawn from verified sources (including TACO-Verified, PrimeIntellect SYNTHETIC-1, and LiveCodeBench submissions), each problem requiring a verifiable solution and at least five unit tests to ensure reliability for RL training. To handle long-range context, DeepCoder employs techniques such as iterative context lengthening and overlong filtering.Starting Price: Free -
33
Yi-Lightning
Yi-Lightning
Yi-Lightning, developed by 01.AI under the leadership of Kai-Fu Lee, represents the latest advancement in large language models with a focus on high performance and cost-efficiency. It boasts a maximum context length of 16K tokens and is priced at $0.14 per million tokens for both input and output, making it remarkably competitive. Yi-Lightning leverages an enhanced Mixture-of-Experts (MoE) architecture, incorporating fine-grained expert segmentation and advanced routing strategies, which contribute to its efficiency in training and inference. This model has excelled in various domains, achieving top rankings in categories like Chinese, math, coding, and hard prompts on the chatbot arena, where it secured the 6th position overall and 9th in style control. Its development included comprehensive pre-training, supervised fine-tuning, and reinforcement learning from human feedback, ensuring both performance and safety, with optimizations in memory usage and inference speed. -
34
DeepEval
Confident AI
DeepEval is a simple-to-use, open source LLM evaluation framework, for evaluating and testing large-language model systems. It is similar to Pytest but specialized for unit testing LLM outputs. DeepEval incorporates the latest research to evaluate LLM outputs based on metrics such as G-Eval, hallucination, answer relevancy, RAGAS, etc., which uses LLMs and various other NLP models that run locally on your machine for evaluation. Whether your application is implemented via RAG or fine-tuning, LangChain, or LlamaIndex, DeepEval has you covered. With it, you can easily determine the optimal hyperparameters to improve your RAG pipeline, prevent prompt drifting, or even transition from OpenAI to hosting your own Llama2 with confidence. The framework supports synthetic dataset generation with advanced evolution techniques and integrates seamlessly with popular frameworks, allowing for efficient benchmarking and optimization of LLM systems.Starting Price: Free -
35
Entry Point AI
Entry Point AI
Entry Point AI is the modern AI optimization platform for proprietary and open source language models. Manage prompts, fine-tunes, and evals all in one place. When you reach the limits of prompt engineering, it’s time to fine-tune a model, and we make it easy. Fine-tuning is showing a model how to behave, not telling. It works together with prompt engineering and retrieval-augmented generation (RAG) to leverage the full potential of AI models. Fine-tuning can help you to get better quality from your prompts. Think of it like an upgrade to few-shot learning that bakes the examples into the model itself. For simpler tasks, you can train a lighter model to perform at or above the level of a higher-quality model, greatly reducing latency and cost. Train your model not to respond in certain ways to users, for safety, to protect your brand, and to get the formatting right. Cover edge cases and steer model behavior by adding examples to your dataset.Starting Price: $49 per month -
36
CodeGemma
Google
CodeGemma is a collection of powerful, lightweight models that can perform a variety of coding tasks like fill-in-the-middle code completion, code generation, natural language understanding, mathematical reasoning, and instruction following. CodeGemma has 3 model variants, a 7B pre-trained variant that specializes in code completion and generation from code prefixes and/or suffixes, a 7B instruction-tuned variant for natural language-to-code chat and instruction following; and a state-of-the-art 2B pre-trained variant that provides up to 2x faster code completion. Complete lines, and functions, and even generate entire blocks of code, whether you're working locally or using Google Cloud resources. Trained on 500 billion tokens of primarily English language data from web documents, mathematics, and code, CodeGemma models generate code that's not only more syntactically correct but also semantically meaningful, reducing errors and debugging time. -
37
Unsloth
Unsloth
Unsloth is an open source platform designed to accelerate and optimize the fine-tuning and training of Large Language Models (LLMs). It enables users to train custom models, such as ChatGPT, in just 24 hours instead of the typical 30 days, achieving speeds up to 30 times faster than Flash Attention 2 (FA2) while using 90% less memory. Unsloth supports both LoRA and QLoRA fine-tuning techniques, allowing for efficient customization of models like Mistral, Gemma, and Llama versions 1, 2, and 3. Unsloth's efficiency stems from manually deriving computationally intensive mathematical steps and handwriting GPU kernels, resulting in significant performance gains without requiring hardware modifications. Unsloth delivers a 10x speed increase on a single GPU and up to 32x on multi-GPU systems compared to FA2, with compatibility across NVIDIA GPUs from Tesla T4 to H100, and portability to AMD and Intel GPUs.Starting Price: Free -
38
FinetuneDB
FinetuneDB
Capture production data, evaluate outputs collaboratively, and fine-tune your LLM's performance. Know exactly what goes on in production with an in-depth log overview. Collaborate with product managers, domain experts and engineers to build reliable model outputs. Track AI metrics such as speed, quality scores, and token usage. Copilot automates evaluations and model improvements for your use case. Create, manage, and optimize prompts to achieve precise and relevant interactions between users and AI models. Compare foundation models, and fine-tuned versions to improve prompt performance and save tokens. Collaborate with your team to build a proprietary fine-tuning dataset for your AI models. Build custom fine-tuning datasets to optimize model performance for specific use cases. -
39
Solar Mini
Upstage AI
Solar Mini is a pre‑trained large language model that delivers GPT‑3.5‑comparable responses with 2.5× faster inference while staying under 30 billion parameters. It achieved first place on the Hugging Face Open LLM Leaderboard in December 2023 by combining a 32‑layer Llama 2 architecture, initialized with high‑quality Mistral 7B weights, with an innovative “depth up‑scaling” (DUS) approach that deepens the model efficiently without adding complex modules. After DUS, continued pretraining restores and enhances performance, and instruction tuning in a QA format, especially for Korean, refines its ability to follow user prompts, while alignment tuning ensures its outputs meet human or advanced AI preferences. Solar Mini outperforms competitors such as Llama 2, Mistral 7B, Ko‑Alpaca, and KULLM across a variety of benchmarks, proving that compact size need not sacrifice capability.Starting Price: $0.1 per 1M tokens -
40
Qwen3-Coder
Qwen
Qwen3‑Coder is an agentic code model available in multiple sizes, led by the 480B‑parameter Mixture‑of‑Experts variant (35B active) that natively supports 256K‑token contexts (extendable to 1M) and achieves state‑of‑the‑art results comparable to Claude Sonnet 4. Pre‑training on 7.5T tokens (70 % code) and synthetic data cleaned via Qwen2.5‑Coder optimized both coding proficiency and general abilities, while post‑training employs large‑scale, execution‑driven reinforcement learning, scaling test‑case generation for diverse coding challenges, and long‑horizon RL across 20,000 parallel environments to excel on multi‑turn software‑engineering benchmarks like SWE‑Bench Verified without test‑time scaling. Alongside the model, the open source Qwen Code CLI (forked from Gemini Code) unleashes Qwen3‑Coder in agentic workflows with customized prompts, function calling protocols, and seamless integration with Node.js, OpenAI SDKs, and environment variables.Starting Price: Free -
41
Amazon Nova Micro
Amazon
Amazon Nova Micro is an AI model designed for high-speed, low-cost text processing and generation. It excels in language understanding, translation, code completion, and mathematical problem-solving, providing fast responses with a generation speed of over 200 tokens per second. The model supports fine-tuning for text input and is ideal for applications requiring real-time processing and efficiency. With support for 200+ languages and a maximum of 128k tokens, Nova Micro is perfect for interactive AI applications that prioritize speed and affordability. -
42
Z-Image
Z-Image
Z-Image is an open source image generation foundation model family developed by Alibaba’s Tongyi-MAI team that uses a Scalable Single-Stream Diffusion Transformer architecture to generate photorealistic and creative images from text prompts with only 6 billion parameters, making it more efficient than many larger models while still delivering competitive quality and instruction following. It includes multiple variants; Z-Image-Turbo, a distilled version optimized for ultra-fast inference with as few as eight function evaluations and sub-second generation on appropriate GPUs; Z-Image, the full foundation model suited for high-fidelity creative generation and fine-tuning; Z-Image-Omni-Base, a versatile base checkpoint for community-driven development; and Z-Image-Edit, tuned for image-to-image editing tasks with strong instruction adherence.Starting Price: Free -
43
Llama Guard
Meta
Llama Guard is an open-source safeguard model developed by Meta AI to enhance the safety of large language models in human-AI conversations. It functions as an input-output filter, classifying both prompts and responses into safety risk categories, including toxicity, hate speech, and hallucinations. Trained on a curated dataset, Llama Guard achieves performance on par with or exceeding existing moderation tools like OpenAI's Moderation API and ToxicChat. Its instruction-tuned architecture allows for customization, enabling developers to adapt its taxonomy and output formats to specific use cases. Llama Guard is part of Meta's broader "Purple Llama" initiative, which combines offensive and defensive security strategies to responsibly deploy generative AI models. The model weights are publicly available, encouraging further research and adaptation to meet evolving AI safety needs. -
44
Dolly
Databricks
Dolly is a cheap-to-build LLM that exhibits a surprising degree of the instruction following capabilities exhibited by ChatGPT. Whereas the work from the Alpaca team showed that state-of-the-art models could be coaxed into high quality instruction-following behavior, we find that even years-old open source models with much earlier architectures exhibit striking behaviors when fine tuned on a small corpus of instruction training data. Dolly works by taking an existing open source 6 billion parameter model from EleutherAI and modifying it ever so slightly to elicit instruction following capabilities such as brainstorming and text generation not present in the original model, using data from Alpaca.Starting Price: Free -
45
Athene-V2
Nexusflow
Athene-V2 is Nexusflow's latest 72-billion-parameter model suite, fine-tuned from Qwen 2.5 72B, designed to compete with GPT-4o across key capabilities. This suite includes Athene-V2-Chat-72B, a state-of-the-art chat model that matches GPT-4o in multiple benchmarks, excelling in chat helpfulness (Arena-Hard), code completion (ranking #2 on bigcode-bench-hard), mathematics (MATH), and precise long log extraction. Additionally, Athene-V2-Agent-72B balances chat and agent functionalities, offering concise, directive responses and surpassing GPT-4o in Nexus-V2 function calling benchmarks focused on complex enterprise-level use cases. These advancements underscore the industry's shift from merely scaling model sizes to specialized customization, illustrating how targeted post-training processes can finely optimize models for distinct skills and applications. -
46
Llama 3.2
Meta
The open-source AI model you can fine-tune, distill and deploy anywhere is now available in more versions. Choose from 1B, 3B, 11B or 90B, or continue building with Llama 3.1. Llama 3.2 is a collection of large language models (LLMs) pretrained and fine-tuned in 1B and 3B sizes that are multilingual text only, and 11B and 90B sizes that take both text and image inputs and output text. Develop highly performative and efficient applications from our latest release. Use our 1B or 3B models for on device applications such as summarizing a discussion from your phone or calling on-device tools like calendar. Use our 11B or 90B models for image use cases such as transforming an existing image into something new or getting more information from an image of your surroundings.Starting Price: Free -
47
Reka
Reka
Our enterprise-grade multimodal assistant carefully designed with privacy, security, and efficiency in mind. We train Yasa to read text, images, videos, and tabular data, with more modalities to come. Use it to generate ideas for creative tasks, get answers to basic questions, or derive insights from your internal data. Generate, train, compress, or deploy on-premise with a few simple commands. Use our proprietary algorithms to personalize our model to your data and use cases. We design proprietary algorithms involving retrieval, fine-tuning, self-supervised instruction tuning, and reinforcement learning to tune our model on your datasets. -
48
Smaug-72B
Abacus
Smaug-72B is a powerful open-source large language model (LLM) known for several key features: High Performance: It currently holds the top spot on the Hugging Face Open LLM leaderboard, surpassing models like GPT-3.5 in various benchmarks. This means it excels at tasks like understanding, responding to, and generating human-like text. Open Source: Unlike many other advanced LLMs, Smaug-72B is freely available for anyone to use and modify, fostering collaboration and innovation in the AI community. Focus on Reasoning and Math: It specifically shines in handling reasoning and mathematical tasks, attributing this strength to unique fine-tuning techniques developed by Abacus AI, the creators of Smaug-72B. Based on Qwen-72B: It's technically a fine-tuned version of another powerful LLM called Qwen-72B, released by Alibaba, further improving upon its capabilities. Overall, Smaug-72B represents a significant step forward in open-source AI.Starting Price: Free -
49
Devstral
Mistral AI
Devstral is an open source, agentic large language model (LLM) developed by Mistral AI in collaboration with All Hands AI, specifically designed for software engineering tasks. It excels at navigating complex codebases, editing multiple files, and resolving real-world issues, outperforming all open source models on the SWE-Bench Verified benchmark with a score of 46.8%. Devstral is fine-tuned from Mistral-Small-3.1 and features a long context window of up to 128,000 tokens. It is optimized for local deployment on high-end hardware, such as a Mac with 32GB RAM or an Nvidia RTX 4090 GPU, and is compatible with inference frameworks like vLLM, Transformers, and Ollama. Released under the Apache 2.0 license, Devstral is available for free and can be accessed via Hugging Face, Ollama, Kaggle, Unsloth, and LM Studio.Starting Price: $0.1 per million input tokens -
50
FLUX.1 Krea
Krea
FLUX.1 Krea is an open source, guidance-distilled 12 billion-parameter diffusion transformer released by Krea in collaboration with Black Forest Labs, engineered to deliver superior aesthetic control and photorealism while eschewing the generic “AI look.” Fully compatible with the FLUX.1-dev ecosystem, it starts from a raw, untainted base model (flux-dev-raw) rich in world knowledge and employs a two-phase post-training pipeline, supervised fine-tuning on a hand-curated mix of high-quality and synthetic samples, followed by reinforcement learning from human feedback using opinionated preference data, to bias outputs toward a distinct style. By leveraging negative prompts during pre-training, custom loss functions for classifier-free guidance, and targeted preference labels, it achieves significant quality improvements with under one million examples, all without extensive prompting or additional LoRA modules.Starting Price: Free
