Skip to main content

Loading Tavory AI

Loading Tavory AI

HomeProvidersNovita

Provider directory

Novita models in Tavory

Tavory currently lists 37 Novita models in its live catalog. Explore model capabilities, catalog signals and eligible workspace access. Tavory does not represent a manufacturer relationship unless explicitly stated.

Open Tavory workspace

Models from Novita

text

Deepseek R1

DeepSeek R1 is the latest open-source model released by the DeepSeek team, featuring impressive reasoning capabilities, particularly achieving performance comparable to OpenAI's o1 model in mathematics, coding, and reasoning tasks.

text

Kat Coder Pro

KAT-Coder-Pro V2 by KwaiKAT is a non-reasoning model optimized for agentic coding. It delivers strong performance on reasoning-style tasks while requiring significantly fewer output tokens than peer models. With the 1210 release, it achieved a score of 64 on the Artificial Analysis Intelligence Index, placing it in the global Top 10 and ranking first among all non-reasoning models.

text

Wizardlm 2 8x22b

WizardLM-2 8x22B is Microsoft AI's most advanced Wizard model. It demonstrates highly competitive performance compared to leading proprietary models, and it consistently outperforms all existing state-of-the-art opensource models.

text

Qwen3.8 Max

A 2.4T-parameter MoE flagship for coding and knowledge work. Programs autonomously for days to deliver complete projects end to end, with native visual understanding across planning, execution, and verification, plus deep semantic analysis of ultra-long documents and video.

text

Kimi K2 Instruct

Kimi K2 Instruct

text

Deepseek V4 Flash

DeepSeek-V4-Flash is a lightweight model meticulously designed by DeepSeek to deliver the ultimate combination of lightning-fast response times and unmatched cost-effectiveness. Engineered with fewer parameters and significantly lower activation overhead, V4-Flash provides an exceptionally fast and economical API service. At its core, V4-Flash demonstrates outstanding reasoning capabilities that closely rival the V4-Pro model. While featuring a slightly streamlined repository of world knowledge,

text

Deepseek R1 Turbo

DeepSeek R1 is the latest open-source model released by the DeepSeek team, featuring impressive reasoning capabilities, particularly achieving performance comparable to OpenAI's o1 model in mathematics, coding, and reasoning tasks.

text

Llama 3.3 70b Instruct

The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks. Supported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.

text

Minimax M2.7

MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement. Built to actively participate in its own evolution, M2.7 integrates advanced agentic capabilities through multi-agent collaboration, enabling it to plan, execute, and refine complex tasks across dynamic environments. Trained for production-grade performance, M2.7 handles workflows such as live debugging, root cause analysis, financial modeling, and full document generation across Word, Excel, and PowerPoint. It delivers strong results on benchmarks including 56.2% on SWE-Pro and 57.0% on Terminal Bench 2, while achieving a 1495 ELO on GDPval-AA, setting a new standard for multi-agent systems operating in real-world digital workflows.

text

L31 70b Euryale V2.2

Euryale L3.1 70B v2.2 is a model focused on creative roleplay from Sao10k. It is the successor of Euryale L3 70B v2.1.

text

Glm 4.6

GLM-4.6 is Z AI’s latest flagship model, designed to push agentic and coding performance further. It expands the context window from 128K to 200K tokens, improves reasoning and tool-use capabilities, and delivers stronger results in coding benchmarks and real-world development workflows. GLM-4.6 demonstrates refined writing quality, more capable agent behavior, and higher token efficiency (≈15% fewer tokens vs. GLM-4.5). Evaluations show clear gains over GLM-4.5 across reasoning, agents, and coding, reaching near parity with Claude Sonnet 4 in practical tasks while outperforming other open-source baselines. GLM-4.6 is available through the Z.ai API platform, OpenRouter, coding agents (Claude Code, Roo Code, Cline, Kilo Code), and soon as downloadable weights on HuggingFace and ModelScope.

text

Deepseek V4 Flash 0731

DeepSeek V4 Flash 0731 is the official release of DeepSeek V4 Flash, superseding the preview version, with substantially enhanced agentic capabilities. A sparse MoE model with 13B active parameters out of 284B total, 1M token context window, suited for coding, reasoning, and agent workflows with function calling and structured output support.

text

Hy3

Hy3 is built for real world business scenarios with a 295B total, 21B active MoE architecture, native 256K context support, and three reasoning modes. It enhances coding, long form comprehension, multi turn dialogue, and agentic task execution, balancing reliability, efficiency, and cost across high frequency interactions and complex workflows.

text

Ring 2.6 1t

Inclusion AI ring-2.6-1t

text

Ling 2.6 Flash

Inclusion AI ling-2.6-flash

text

Glm 5.1

GLM-5.1 is Z.AI's latest flagship model, meticulously engineered for long-horizon autonomous tasks. Capable of working continuously on a single assignment for up to 8 hours, it autonomously manages the entire workflow—from initial planning and execution to iterative optimization and the delivery of production-grade results. While its general capabilities and coding proficiency are on par with Claude Opus 4.6, GLM-5.1 excels particularly in sustained execution, complex engineering optimization, a

text

Deepseek V3

DeepSeek-V3 is the latest model from the DeepSeek team, building upon the instruction following and coding abilities of the previous versions. Pre-trained on nearly 15 trillion tokens, the reported evaluations reveal that the model outperforms other open-source models and rivals leading closed-source models.

text

Llama 3.1 8b Instruct

Meta's latest class of models, Llama 3.1, launched with a variety of sizes and configurations. The 8B instruct-tuned version is particularly fast and efficient. It has demonstrated strong performance in human evaluations, outperforming several leading closed-source models.

text

Qwen 2.5 72b Instruct

Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters.

video

Qwen3.5 397b A17b

The Qwen3.5 series 397B-A17B native vision-language model is based on a hybrid architecture design that integrates linear attention mechanisms with sparse Mixture-of-Experts (MoE), achieving higher inference efficiency. Across a variety of tasks—including language understanding, logical reasoning, code generation, agentic tasks, image understanding, video understanding, and graphical user interface (GUI) interaction—it demonstrates exceptional performance comparable to current top-tier frontier models. Possessing robust code generation and agentic capabilities, it exhibits strong generalization across various agent scenarios.

text

Ling 3.0 Tiny

Ling-3.0-tiny is an efficient 7.9B parameter MoE model from inclusionAI with only 1.3B active parameters per token. Built for responsive agents, reliable instruction following and multi turn conversation, with a 256K context window, native function calling, prompt caching and switchable Thinking and Instant modes.

text

Deepseek R1 Distill Llama 70b

DeepSeek R1 Distill LLama 70B

text

Minimax M2.7 Highspeed

MiniMax M2.7-highspeed is an accelerated SOTA model engineered for scenarios demanding extreme efficiency. It perfectly inherits the core intelligence and robust digital workspace capabilities of the standard M2.7. In real-world software engineering, M2.7 excels by independently driving end-to-end project delivery while efficiently handling advanced tasks such as log analysis, bug troubleshooting, code security, and machine learning. In the professional workspace, it boasts the highest open-sour

text

Mythomax L2 13b

The idea behind this merge is that each layer is composed of several tensors, which are in turn responsible for specific functions. Using MythoLogic-L2's robust understanding as its input and Huginn's extensive writing capability as its output seems to have resulted in a model that exceeds at both, confirming my theory. (More details to be released at a later time).

text

Deepseek V3 Turbo

DeepSeek R1 is the latest open-source model released by the DeepSeek team, featuring impressive reasoning capabilities, particularly achieving performance comparable to OpenAI's o1 model in mathematics, coding, and reasoning tasks.

text

Deepseek V3 0324

DeepSeek R1 is the latest open-source model released by the DeepSeek team, featuring impressive reasoning capabilities, particularly achieving performance comparable to OpenAI's o1 model in mathematics, coding, and reasoning tasks.

text

Qwen3 235b A22b Fp8

Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance. Significantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.

text

Mistral Nemo

A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA. The model is multilingual, supporting English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi. It supports function calling and is released under the Apache 2.0 license.

text

L3 8b Lunaris

A generalist / roleplaying model merge based on Llama 3.

text

Deepseek V3.2

DeepSeek-V3.2 is a next-generation foundation model designed to unify high computational efficiency with state-of-the-art reasoning and agentic performance. Built upon DeepSeek Sparse Attention (DSA) for efficient long-context reasoning, a scalable reinforcement learning framework reaching frontier-level performance, and a large-scale agentic task synthesis pipeline for reliable tool-use and multi-step decision-making.

text

Llama 4 Maverick 17b 128e Instruct Fp8

A lightweight and ultra-fast variant of Llama 3.3 70B, for use when quick response times are needed most.

text

Ling 2.6 1t

Inclusion AI ling-2.6-1t

text

Baichuan M2 32b

Baichuan-M2 is a medically-enhanced reasoning model specifically designed for real-world medical reasoning tasks. We begin with real-world medical questions and conduct reinforcement learning training based on a large-scale verifier system. While maintaining the model's general capabilities, the medical effectiveness of Baichuan-M2 has achieved breakthrough improvements. Baichuan-M2 is currently the world's best open-source medical model. On the HealthBench Benchmark, it surpasses all open-sour

text

Step 3.7 Flash

Step 3.7 Flash is a fast reasoning model from StepFun with a 256K context window, supporting function calling and structured output for high volume agentic workloads.

text

Gemma 4 26b A4b It

Gemma 4 26B A4B is a Mixture of Experts model from Google DeepMind built for scalable performance with a 256K context window. Supports text and image input, native thinking mode, function calling, structured output, and 140+ languages. Ideal for long context RAG, agentic workflows, coding, and multimodal understanding.

text

L3 8b Stheno V3.2

Sao10K/L3-8B-Stheno-v3.2 is a highly skilled actor that excels at fully immersing itself in any role assigned.

text

GLM 5

GLM-5 is an open-source foundation model engineered for complex system engineering and long-horizon Agent tasks, delivering reliable productivity for top-tier programmers. Transcending the boundary from "writing code" to "building systems," it moves beyond traditional snippet generation to offer senior-architect-level planning and execution capabilities. By rejecting the "frontend-heavy, logic-light" approach, GLM-5 demonstrates exceptional reasoning and self-healing abilities in backend refactoring, complex algorithm implementation, and deep debugging—autonomously analyzing logs and iteratively fixing persistent bugs until the system runs. As the first open-source model featuring Opus-class style and system engineering depth, GLM-5 provides extreme logic density alongside the freedom of local deployment and high cost-effectiveness, making it the ideal choice for large-scale backend development and automated Agent construction.

Using Novita models in Tavory

Choose an eligible model from the catalog or let Smart Mode select a suitable available route. Tavory checks capabilities, plan access and current availability server-side when a request is sent.

Open Tavory workspace