API

API keys

Generate up to 5 API keys to use NanoGPT in other applications. If you require more keys, please contact us at support@nano-gpt.com and we will help you out.

Authenticate by including your API key as a HTTP header: "Authorization": f"Bearer API_KEY" or "api-key": "API_KEY" depending on the endpoint.

NameStatusCreatedAPI Key

Get notified about API updates.

We will only use this to contact you updates to how the API works. You can unsubscribe at any time.

If you are a (potentially) large user of our website or our API, we are glad to have you. Reach out to us at support@nano-gpt.com or join our Discord for a discount.

API Reference

The below example code can be used in Python, NanoGPTjs is a great starting point for JS users.

If you encounter issues or need further information please contact support@nano-gpt.com

Text models
POST https://nano-gpt.com/api/talk-to-gpt
NameModelDescription
Auto modelrecommended-modelAutomatically uses the best model for your task. Categorizes the prompt, then uses the model that performs best in that particular category according to global user preferences. Scores updated daily. Ability to set pricing tier in Adjust Settings.
Grok 3 Thinkinggrok-3-reasonerGrok 3 Thinking adds chain of thought into the Grok 3 model leading to even better results across a wide range of tasks. Displays its thinking. Note: currently heavily rate limited.
Grok 3grok-3Grok 3 is the newest xAI model and current leader of most leaderboards. Comes with a massive 1 million token context window.
Grok 3 Deepsearchgrok-3-deepsearchGrok 3 Deepsearch is a lightning-fast AI agent built to relentlessly seek the truth across the entire corpus of human knowledge. DeepSearch is designed to synthesize key information, reason about conflicting facts and opinions, and distill clarity from complexity. Note: currently heavily rate limited.
OpenAI o3-minio3-miniOpenAI's newest flagship model.
OpenAI o3-mini higho3-mini-highOpenAI's newest flagship model with reasoning effort set to high.
OpenAI o3-mini lowo3-mini-lowOpenAI's newest flagship model with reasoning effort set to low.
OpenAI o1 Proo1-plusNote: Degraded service. Not fully stable. Occasionally fails to respond. The Pro version of OpenAI's flagship reasoning model for solving hard problems. Useful when tackling complex problems in science, coding, math, and similar fields. Minimum cost is ~$0.20, we are temporarily charging max $0.50 regardless of how big your prompt is. This model can think for a long time - please be patient.
OpenAI o1o1OpenAI's flagship reasoning model for solving hard problems. Useful when tackling complex problems in science, coding, math, and similar fields.
ChatGPT 4ochatgpt-4o-latestOpenAI's current recommended model, the well-known ChatGPT.
Gemini 2.0 Pro Exp 0205gemini-2.0-pro-exp-02-05Gemini 2.0 Pro Experimental, the latest version of the Gemini 2.0 Pro model.
Gemini 2.0 Pro Exp 1206gemini-exp-1206Gemini 2.0 Pro Experimental, the latest version of the Gemini 2.0 Pro model.
Model Recommendermodel-selectorModel Recommender - input your query to have it recommend the best model for your task
Kimi K1.5 Previewkimi-k1.5-previewKimi K1.5 is an o1-level multimodal model by Moonshot AI, outperforming o1 on many benchmarks.
Gemini 2.0 Flash Thinking 0121gemini-2.0-flash-thinking-exp-01-21Google's newest model, outperforming even Gemini 1.5 Pro, now with a thinking mode enabled similar to the o1 series of OpenAI.
Gemini 2.0 Flash Thinking 1219gemini-2.0-flash-thinking-exp-1219Google's newest model, outperforming even Gemini 1.5 Pro, now with a thinking mode enabled similar to the o1 series of OpenAI.
Gemini 2.0 Flash Expgemini-2.0-flash-exp-searchGoogle's newest model, outperforming even Gemini 1.5 Pro. Now with web access.
Gemini 2.0 Flash Expgemini-2.0-flash-expGoogle's newest model, outperforming even Gemini 1.5 Pro.
OpenAI o1 previewo1-previewOpenAI's new flagship series of reasoning models for solving hard problems. Useful when tackling complex problems in science, coding, math, and similar fields
OpenAI o1-minio1-miniA fast, cost-efficient version of OpenAI's o1 reasoning model tailored to coding, math, and science use cases.
Open Deep Researchdeep-researcho3-mini-powered research assistant that performs deep analysis across multiple sources. Open source version of OpenAI's Deep Research. Warning: thinks for a while - it has to write an entire report! Based on open-deep-research by fdarkaou.
Grok 2 1212grok-2-1212Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model..
Grok 2 1212x-ai/grok-2-1212Grok 2 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model..
Doubao 1.5 Pro 256kdoubao-1.5-pro-256kDoubao's (Bytedance) flagship model with a 256k token context window
Gemini 2.0 Flashgemini-2.0-flash-001Upgraded version of Gemini Flash 1.5. Faster, with higher output, and overall increase in intelligence.
Gemini 2.0 Flash Lite Previewgemini-2.0-flash-lite-preview-02-05Upgraded version of Gemini Flash 1.5. Faster, with higher output, and overall increase in intelligence.
Grok 2 Betagrok-betaGrok-2 is xAI's frontier language model, the one used on X. Claims state-of-the-art reasoning capabilities, best for complex and multi-step use cases.
Grok 2 Betax-ai/grok-betaGrok-2 is xAI's frontier language model, the one used on X. Claims state-of-the-art reasoning capabilities, best for complex and multi-step use cases.
Claude 3.5 Sonnetclaude-3-5-sonnet-20241022Anthropic's updated most intelligent model, offering even better results on many subjects than GPT-4o.
DeepSeek R1deepseek-r1-nanoDeepSeek's R1 is a thinking model, rivalling OpenAI's o1. This version is run via Azure, with fallbacks to Azure, Fireworks and Together, never routing through DeepSeek themselves.
Aion 1.0 mini (DeepSeek)aion-labs/aion-1.0-miniA distilled version of the DeepSeek-R1 model that excels in reasoning domains like mathematics, coding, and logic.
Aion 1.0aion-labs/aion-1.0Aion Labs most powerful reasoning model with high performance across reasoning and coding.
DeepClaudedeepclaudeHarness the power of DeepSeek R1's reasoning combined with Claude's creativity and code generation. Feeds your query into DeepSeek R1, then feeds the query + thinking process into Claude 3.5 Sonnet and returns an answer. Note: this routes through original DeepSeek meaning your data may be stored and used by DeepSeek.
DeepSeek V3/Deepseek Chatdeepseek-chatLatest model from DeepSeek, trained on nearly 15 trillion tokens, matches leading closed-source models at a far lower price.
DeepSeek V3/Deepseek Chatdeepseek/deepseek-chatLatest model from DeepSeek, trained on nearly 15 trillion tokens, matches leading closed-source models at a far lower price.
DeepSeek R1 Llama 70bdeepseek-r1-llama-70bDeepSeek R1 Llama 70b is a fine-tuned version of DeepSeek R1 on Llama 70B.
MiniMax 01minimax/minimax-01MiniMax's flagship model with a 1M token context window
MiniMax 01MiniMax-Text-01MiniMax's flagship model with a 1M token context window
GLM Zero Previewglm-zero-previewGLM Zero Preview is a thinking model like o1, but with a smaller context window
Step-2 16k Expstep-2-16k-expStep-2 16k Exp is a 16k context window model
GLM 4 Plus 0111glm-4-plus-0111GLM 4 Plus 0111 is a 1M token context window model
GLM 4 Air 0111glm-4-air-0111MiniMax's flagship model with a 1M token context window
Step-2 Ministep-2-miniMiniMax's flagship model with a 1M token context window
Doubao 1.5 Pro 32kdoubao-1.5-pro-32kDoubao's (Bytedance) pro model with a 32k token context window
Doubao 1.5 Vision Pro 32kdoubao-1.5-vision-pro-32kDoubao's (Bytedance) vision-enabled pro model (JPG only) with a 32k token context window
Qwen QwQ 32B PreviewQwen/QwQ-32B-PreviewExperimental release of Qwen's reasoning model. Great at coding and math, but still in development so may exhibit odd bugs. Not production-ready.
Kimi Latestkimi-latestAlways point to the latest stable Kimi model.
Step-2 16kstep-2-16kChinese-based trillion-parameter model by StepFun that scores extremely well on Livebench for a broad range of tasks. Supports a variety of languages, but has a relatively small context window (~8000 words).
Llama 3.3 70b Instructllama-3.3-70bLlama 3.3 is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.
Llama 3.3 70b Instructmeta-llama/llama-3.3-70b-instructLlama 3.3 is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.
Dolphin 72bdolphin-2.9.2-qwen2-72bDolphin is the most uncensored model yet, built on top of Qwen's 72b model.
Nvidia Nemotron 70bnvidia/Llama-3.1-Nemotron-70B-Instruct-HFNvidia's latest Llama fine-tune optimized for instruction following. Early results hints that it might outperform models such as GPT-4o and Claude 3.5 Sonnet.
Claude 3.5 Sonnet Oldclaude-3-5-sonnet-20240620Anthropic's most intelligent model, offering even better results on many subjects than GPT-4o.
Claude 3.5 Haikuclaude-3-5-haiku-20241022Anthropic's updated faster and cheaper model, offering good results on chatbots and coding.
Claude 3 Opusclaude-3-opus-20240229Anthropic's flagship model, outperforming GPT-4 on most benchmarks.
Yi Lightningyi-lightningChinese-developed multilingual (English, Chinese and others) model by 01.ai that's very fast and cheap, yet scores high on independent leaderboards.
Amazon Nova Pro 1.0amazon/nova-pro-v1Amazon's new flagship model. Can handle up to 300k input tokens, with comparable performance to ChatGPT and Claude 3.5 Sonnet.
GPT 4o minigpt-4o-miniOpenAI's most cost-efficient small model. Cheaper and smarter than GPT-3.5 (the original ChatGPT), but less performant than gpt-4o
GLM-4 Plusglm-4-plusGLM high-intelligence flagship model with 128K context window
Gemini LearnLM Experimentallearnlm-1.5-pro-experimentalLearnLM is a task-specific model trained to align with learning science principles when following system instructions for teaching and learning use cases. For instance, the model can take on tasks to act as an expert or guide to educate users on specific topics.
Llama 3.1 LargeMeta-Llama-3-1-405B-Instruct-FP8Note: comes with a 90% discount currently, enjoy! Meta's largest Llama 3.1 405B model. Open-source, run through an open permissionless crypto network (no central provider).
Hermes 3 Largenousresearch/hermes-3-llama-3.1-405bLlama 3.1 405b with the brakes taken off. Less censored than the regular version, but not abliterated
Qwen Turboqwen-turboAlibaba's fastest and cheapest model. Suitable for simple tasks, fast and low cost, with a 1 million token context window.
Qwen 2.5 Maxqwen-maxQwen 2.5 Max is the upgraded version of Qwen Max, beating GPT-4o, Deepseek V3 and Claude 3.5 Sonnet in benchmarks.
Qwen Plusqwen-plusAlibaba's balanced model. Fast, cheap, yet still very powerful.
Qwen Long 10Mqwen-longAlibaba's huge context window model. Takes in up to 10 million tokens, which is equivalent to dozens of books.
Qwen 2.5 Coder 32bQwen/Qwen2.5-Coder-32B-InstructThe latest series of Code-Specific Qwen large language models.
Amazon Nova Lite 1.0amazon/nova-lite-v1Amazon's new lower cost model. Can handle up to 300k input tokens, with faster output but less thorough understanding than Amazon's Nova Pro.
Amazon Nova Micro 1.0amazon/nova-micro-v1Amazon's lowest cost model. Comparable to GPT-4o-mini and Gemini 1.5 Flash, with the fastest output.
Yi Largeyi-largeLarge version of Yi Lightning with a 32k context window, but more expensive.
Yi Medium 200kyi-medium-200kMedium version of Yi with a 200k context window.
Nemo Arli 12b RPMa V1.2Mistral-Nemo-12B-ArliAI-RPMax-v1.2A Mistral Nemo 12b finetuned for roleplay and storytelling.
LatitudeGames WayFarer 12BMistral-Nemo-12B-WayfarerLatitude Games Wayfarer 12B
The Drummer Cydonia 24BTheDrummer/Cydonia-24B-v2Cydonia 24B v2 is a finetune of Mistral's latest 'Small' model (2501). Aliases: Cydonia 24B, Cydonia v2, Cydonia on that broken base.
EVA Llama 3.33 70BEVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0A RP/storywriting specialist model, full-parameter finetune of Llama-3.3-70B-Instruct on mixture of synthetic and natural data. It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and flavor of the resulting model.
Llama 3.1 70B Dracarys 2abacusai/Dracarys-72B-InstructLlama 3.1 70b finetune that offers improvements on coding.
Mistral Large 2411mistralai/mistral-largeUpgrade to Mistral's flagship model. It is fluent in English, French, Spanish, German, and Italian, with high grammatical accuracy, with a long context window.
Lumimaid v0.2NeverSleep/Lumimaid-v0.2-70BUpgrade to Llama-3 Lumimaid 70B. A Llama 3.1 70B finetune trained on curated roleplay data. Extremely uncensored and suitable for NSFW.
DeepSeek V3/Chat Cheaperdeepseek-chat-cheaperCheaper version of Deepseek V3/Chat. Note: may be routed through Deepseek itself.
Inflection 3 Piinflection/inflection-3-piA chatbot with emotional intelligence. Has access to recent news, excels in scenarios like customer support and roleplay. Mirrors your conversation style.
Inflection 3 Productivityinflection/inflection-3-productivityOptimized for instruction following. Good at tasks that require precise adherence to provided guidelines. Has access to recent news.
WizardLM-2 8x22Bmicrosoft/wizardlm-2-8x22bMicrosoft's advanced Wizard model. The most popular role-playing model.
SorcererLM 8x22Braifle/sorcererlm-8x22bAdvanced roleplaying model with reasoning and emotional intelligence for engaging interactions, contextual awareness and enhanced narrative depth
Llama 3.1 Largeaccounts/fireworks/models/llama-v3p1-405b-instructMeta's largest and most capable Llama model. Competitive with GPT-4o and Claude 3.5 Sonnet.
GPT 4o 08 06gpt-4o-2024-08-06OpenAI's precusor to ChatGPT-4o. Great on English text and code, with significant improvements on text in non-English languages.
GPT 4o 11 20gpt-4o-2024-11-20OpenAI's precusor to ChatGPT-4o. Great on English text and code, with significant improvements on text in non-English languages.
Llama 3.2 Mediummeta-llama/llama-3.2-90b-vision-instructMedium-size (and capability) version of Meta's newest model (3.2 series).
Llama 3.1 Mediumaccounts/fireworks/models/llama-v3p1-70b-instructMeta's updated version of their medium Llama model. Slightly lesser performance than Llama Large, but cheaper.
Llama 3.3 70B Instruct abliteratedhuihui-ai/Llama-3.3-70B-Instruct-abliteratedAn abliterated (removed restrictions and censorship) version of Llama 3.3 70b.
Perplexity Prosonar-proSonar Pro tackles complex questions that need deeper research and provides more sources.
Perplexity Reasoning Prosonar-reasoning-proPerplexity's Sonar Reasoning Pro uses DeepSeek R1's thinking process combined with looking up on the web to tackle complex questions that need deeper research and provides more sources.
Perplexity Reasoningsonar-reasoningPerplexity's Sonar Reasoning uses DeepSeek R1's thinking process combined with looking up on the web to tackle complex questions that need deeper research and provides more sources.
Llama 3.1 Largemeta-llama/llama-3.1-405b-instructMeta's largest and most capable Llama model. Competitive with GPT-4o and Claude 3.5 Sonnet.
Gemini 1.5 Flashgoogle/gemini-flash-1.5Google's fastest multimodal model with great performance for diverse, repetitive tasks and a 2 million words context window.
Perplexity SimplesonarA Perplexity model that gives fast, straightforward answers.
MythoMax 13BGryphe/MythoMax-L2-13bOne of the highest performing and most popular fine-tunes of Llama 2 13B, with rich descriptions and roleplay.
GLM-4glm-4High-intelligence model with 128K context window
GLM-4 Longglm-4-longExtended context model supporting up to 1M tokens
Qwen2.5 72Bqwen/qwen-2.5-72b-instructGreat multilingual support, strong at mathematics and coding, supports roleplay and chatbots.
EVA Qwen2.5 72Beva-unit-01/eva-qwen-2.5-72bFull-parameter finetune of Qwen2.5-72B on mixture of synthetic and natural data. It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and flavor of the resulting model.
Yi Medium 200kyi-34b-chat-200kMedium version of Yi Lightning with a huge 200k context window
Yi Sparkyi-34b-chat-0205Small and powerful, lightweight and fast model. Provides enhanced mathematical operation and code writing capabilities.
Yi Large Turboyi-large-turboSuper cost-effective, excellent performance. Balanced high-precision tuning based on performance, inference speed, and cost.
Dolphin 2.6 Mixtral 8x7bcognitivecomputations/dolphin-mixtral-8x7bDesigned for instruction following, conversational, and coding.
GPT 4 Turbo Previewgpt-4-turbo-previewCan take in the largest messages (up to 300 pages of context), and all round seen as one of the best in class models.
GPT 4 Turbogpt-4-turboCan take in the largest messages (up to 300 pages of context), and all round seen as one of the best in class models.
GPT 4ogpt-4oOpenAI's precusor to ChatGPT-4o. Great on English text and code, with significant improvements on text in non-English languages.
GPT 3.5 Turbogpt-3.5-turboOlder model. Brought ChatGPT to the mainstream, seen as dated nowadays. 90% cheaper than GPT-4-Turbo, recommended for very simple tasks.
Gemini 1.5 Flashgemini-1.5-flash-001Google's fastest multimodal model with great performance for diverse, repetitive tasks and a 1 million context window.
Gemini 1.5 Progemini-1.5-pro-001Google's next-generation model with a breakthrough 1 million context window. Comparable to GPT-4o.
Free modelfree-modelFree model to try out our service with. Currently Llama 3.3 70B, but this might change at any time.
Magnum v4 72Banthracite-org/magnum-v4-72bUpgraded model of Magnum V2 72B. From the creators of Goliath. Aimed at achieving prose quality similar to Claude Opus 3, trained on 55 million tokens of curated Roleplay data.
EVA-Qwen2.5-32B-v0.2EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2A RP/storywriting specialist model, full-parameter finetune of Qwen2.5-32B on mixture of synthetic and natural data. It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and flavor of the resulting model.
DeepSeek R1 Distill 70bdeepseek/deepseek-r1-distill-llama-70bDeepSeek-R1 distilled version on Llama 70B.
Llama 3.1 405B Instructnvidia/Llama-3.1-405B-Instruct-FP8NVIDIA's optimized version of Llama 3.1 405B with FP8 precision.
MN-LooseCannon-12B-v1GalrionSoftworks/MN-LooseCannon-12B-v1Merge of Starcannon and Sao Lyra.
EVA-Qwen2.5-72B-v0.2EVA-UNIT-01/EVA-Qwen2.5-72B-v0.2A RP/storywriting specialist model, full-parameter finetune of Qwen2.5-72B on mixture of synthetic and natural data. It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and flavor of the resulting model.
EVA-LLaMA-3.33-70B-v0.1EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1A RP/storywriting specialist model, full-parameter finetune of Llama-3.3-70B-Instruct on mixture of synthetic and natural data. It uses Celeste 70B 0.1 data mixture, greatly expanding it to improve versatility, creativity and flavor of the resulting model.
Dolphin 2.9.2 Mixtral 8x22Bcognitivecomputations/dolphin-mixtral-8x22bSuccessor to Dolphin 2.6 Mixtral 8x7b. Great for instruction following, conversational, and coding.
Llama 3.1 70b Instructmeta-llama/llama-3.1-70b-instructOptimized for high quality dialogue usecases.
Llama 3.1 8b Instructmeta-llama/llama-3.1-8b-instructFast and efficient for simple purposes.
ReMM SLERP 13Bundi95/remm-slerp-l2-13bA recreation trial of the original MythoMax-L2-B13 but merged with updated models.
Mistral Tinymistralai/mistral-tinyPowered by Mistral-7B-v0.2, best used for large batch processing tasks where cost is a significant factor but reasoning capabilities are not crucial.
Mistral Sabamistralai/mistral-sabaMistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications.
Mistral 7B Instructmistralai/mistral-7b-instructOptimized for speed with decent context length
Llama 3 70b Instructmeta-llama/llama-3-70b-instructOptimized for high quality dialogue usecases.
WizardLM-2 7Bmicrosoft/wizardlm-2-7bFinetune of Mistral 7B Instruct, very fast.
DeepSeek R1 Zero Previewdeepseek-ai/DeepSeek-R1-ZeroPreview version of Deepseek R1, also known as DeepSeek R1 Zero. Deepseek R1 without the supervised finetuning.
Cohere: Command Rcohere/command-r35B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents
Cohere: Command R+cohere/command-r-plus-08-2024104B parameter model that performs conversational language tasks at a higher quality, more reliably, and with a longer context than previous models. It can be used for complex workflows like code generation, retrieval augmented generation (RAG), tool use, and agents
Neural Daredevil 8B abliteratedmlabonne/NeuralDaredevil-8B-abliteratedThe best performing 8B abliterated model according to most benchmarks.
Llama 3 70B abliteratedfailspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5An abliterated (removed restrictions and censorship) version of Llama 3.1 70b.
Nemotron 3.1 70B abliteratedhuihui-ai/Llama-3.1-Nemotron-70B-Instruct-HF-abliteratedAn abliterated (removed restrictions and censorship) version of Llama 3.1 70b Nemotron.
Magnum V2 72Banthracite-org/magnum-v2-72bMagnum V2 72B
Damascus R1.Steelskull/L3.3-Damascus-R1Damascus-R1 builds upon some elements of the Nevoria foundation but represents a significant step forward with a completely custom-made DeepSeek R1 Distill base: Hydroblated-R1-V3. Constructed using the new SCE (Select, Calculate, and Erase) merge method, Damascus-R1 prioritizes stability, intelligence, and enhanced awareness.
Mistral Nemomistralai/Mistral-Nemo-Instruct-240712B parameter model with multilingual support.
DeepSeek Reasonerdeepseek-reasonerDeepSeek-R1 is now live and open source, rivaling OpenAI's Model o1.
Llama 3.1 70B ArliAI RPMax v1.3Llama-3.3+3.1-70B-ArliAI-RPMax-v1.3RPMax are a series of models that are trained on a diverse set of curated creative writing and RP datasets with a focus on variety and deduplication. This model is designed to be highly creative and non-repetitive by making sure no two entries in the dataset have repeated characters or situations, which makes sure the model does not latch on to a certain personality and be capable of understanding and acting appropriately to any characters or situations.
Llama 3.05 Storybreaker Ministral 70bEnvoid/Llama-3.05-NT-Storybreaker-Ministral-70BMuch more inclined to output adult content than its predecessor. Great choice for novelty roleplay scenarios.
Nemotron Tenyxchat Storybreaker 70bEnvoid/Llama-3.05-Nemotron-Tenyxchat-Storybreaker-70BOverall it provides a solid option for RP and creative writing while still functioning as an assistant model, if desired. If used to continue a roleplay it will generally follow the ongoing cadence of the conversation.
Mag Mell R1inflatebot/MN-12B-Mag-Mell-R1Mag Mell demonstrates worldbuilding capabilities unlike any model in its class, comparable to old adventuring models like Tiefighter, and prose that exhibits minimal slop.
Evayale 70b Steelskull/L3.3-MS-Evayale-70BCombination of EVA and Euryale.
Lumimaid 70bNeverSleep/Llama-3-Lumimaid-70B-v0.1Neversleep Llama 3 Lumimaid 70B
MS Evalebis 70bSteelskull/L3.3-MS-Evalebis-70bCombination of EVA, Euryale and Anubis.
Anubis 70B v1TheDrummer/Anubis-70B-v1L3.3 finetune for roleplaying.
Qwen 2.5 32b EVAQwen2.5-32B-EVA-v0.2A Qwen 2.5 32b finetuned for roleplay and storytelling.
Llama 3.3 70b Mirai FanfareLlama-3.3-70B-MiraiFanfareA Llama 3.3 70b finetuned for roleplay and storytelling.
Dazzling Star Aurora 32bQwen2.5-32B-Dazzling-Star-Aurora-32b-v0.0A Qwen 2.5 32b finetuned for roleplay and storytelling.
Gemini 1.5 Progoogle/gemini-pro-1.5Google's next-generation model with a breakthrough 4 million context window. Comparable to GPT-4o.
Gemini 2.0 Flash Searchgemini-2.0-flash-searchGemini 2.0 Flash Search is a version of Gemini 2.0 Flash that has been finetuned for search tasks.
Gemini 2.0 Flash Exp Freegoogle/gemini-2.0-flash-exp:freeSmall model optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization
Llama 3.2 3b Instructmeta-llama/llama-3.2-3b-instructSmall model optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization
Llama 3.1 8B (decentralized)Meta-Llama-3-1-8B-Instruct-FP8Meta's Llama 3.1 8B model via an open permissionless network
GLM-4 AirXglm-4-airxFastest GLM-4 variant with 8K context window
GLM-4 Airglm-4-airHigh-performance model with 128K context window
GLM-4 Flashglm-4-flashExtremely cheap model with 128K context window
Llama 3.1 70B HanamiSao10K/L3.1-70B-Hanami-x1Euryale v2.2-based finetune.
Rocinante 12bTheDrummer/Rocinante-12B-v1.1Designed for engaging storytelling and rich prose. Expanded vocabulary with unique and expressive word choices, enhanced creativity and captivating stories.
Llama 3.3 70B EuryaleSao10K/L3.3-70B-Euryale-v2.3A 70B parameter model from SAO10K based on Llama 3.3 70B, offering high-quality text generation.
Llama 3.1 70B EuryaleSao10K/L3.1-70B-Euryale-v2.2A 70B parameter model from SAO10K based on Llama 3.1 70B, offering high-quality text generation.
UnslopNemo 12b v4TheDrummer/UnslopNemo-12B-v4.1UnslopNemo v4 is the previous version from the creator of Rocinante, designed for adventure writing and role-play scenarios.
Nous Hermes 3 70BNousResearch/Hermes-3-Llama-3.1-70BGeneralist language model including advanced agentic capabilities, much better roleplaying, reasoning, multi-turn conversation, long context coherence, and improvements across the board.
NemoMix 12B UnleashedMarinaraSpaghetti/NemoMix-Unleashed-12BGreat for RP and storytelling.
Mistral Nemo Starcannon 12b v1VongolaChouko/Starcannon-Unleashed-12B-v1.0Mistral Nemo finetine that offers improvements on roleplay.
Llama 3.1 70B Celeste v0.1nothingiisreal/L3.1-70B-Celeste-V0.1-BF16Creative model based on Llama 3.1 70B
Mistral Nemo Inferor 12BInfermatic/MN-12B-Inferor-v0.0Inferor is a merge of top roleplay models, expert on immersive narratives and storytelling.
Claude 3.5 Sonnetanthropic/claude-3.5-sonnetAnthropic's updated most intelligent model, offering even better results on many subjects than GPT-4o.
DeepSeek R1aihubmix-DeepSeek-R1DeepSeek's R1 model, offering even better results on many subjects than GPT-4o.
DeepSeek R1deepseek/deepseek-r1DeepSeek-R1 is now live and open source, rivaling OpenAI's Model o1.
Athene V2 ChatNexusflow/Athene-V2-ChatAn open-weights LLM on-par with GPT-4o across benchmarks.
Deepseek R1 Qwen Abliteratedhuihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliteratedUncensored version of the Deepseek R1 Qwen 32B model
Deepseek R1 Llama 70b Abliteratedhuihui-ai/DeepSeek-R1-Distill-Llama-70B-abliteratedUncensored version of the Deepseek R1 Llama 70B model
DeepSeek R1 671Bdeepseek-r1-671bDeepSeek R1 671B model
Deepseek R1 Cheaperdeepseek-reasoner-cheaperCheaper version of DeepSeek R1. Note: may be routed through Chinese providers.
Deepseek R1 Cheaperark-deepseek-r1-250120Cheaper version of DeepSeek R1. Note: may be routed through Chinese providers.
Llama 3.1 8b (uncensored)aion-labs/aion-rp-llama-3.1-8bThis is a truly uncensored model, trained to excel at roleplaying and creative writing. However, it can also do other things!
Azure o1azure-o1Azure version of OpenAI o1
Azure o3-miniazure-o3-miniAzure version of OpenAI o3-mini
Azure gpt-4oazure-gpt-4oAzure version of OpenAI gpt-4o
Azure gpt-4o-miniazure-gpt-4o-miniAzure version of OpenAI gpt-4o-mini
Azure gpt-4-turboazure-gpt-4-turboAzure version of OpenAI gpt-4-turbo
Llama 3.1 Tulu 3 405BLlama-3.1-Tulu-3-405BTülu 3 405B, a fine tune of Llama 405B that performs better than DeepSeek V3 on SambaNova Cloud. This powerful open-source model, developed by the Allen Institute for AI (Ai2), represents a significant leap forward in large language model capabilities. Thanks to the SambaNova RDU, we are able to efficiently support this model at over 90tokens/second.
DeepSeek R1 Sambanovadeepseek-r1-sambanovaDeepSeek R1 via Sambanova: the full model with very fast output. Note: max 4k output tokens.
Grok 2 Vision 1212grok-2-vision-1212Grok 2 Vision 1212 introduces significant enhancements to accuracy, instruction adherence, and multilingual support, making it a powerful and flexible choice for developers seeking a highly steerable, intelligent model..
DeepSeek V3/Deepseek Chatdeepseek/deepseek-chat:freeDeepSeek Chat is a model that is a good choice for general purpose chat.
DeepSeek R1deepseek/deepseek-r1:freeDeepSeek R1 is a model that is a good choice for general purpose chat.
Perplexity R1 1776r1-1776R1 1776 is a version of the DeepSeek R1 model that has been post-trained by Perplexity to provide uncensored, unbiased, and factual information.
Llama 3.3 70B WayfarerLatitudeGames/Wayfarer-Large-70B-Llama-3.3Llama 3.3 70B Wayfarer is a fine-tuned version of Llama 3.3 70B, trained on a diverse set of creative writing and RP datasets with a focus on variety and deduplication. This model is designed to be highly creative and non-repetitive by making sure no two entries in the dataset have repeated characters or situations, which makes sure the model does not latch on to a certain personality and be capable of understanding and acting appropriately to any characters or situations.
Image models
POST https://nano-gpt.com/api/generate-image
NameModelDescription
Recraft V3recraft-v3The current best scoring model across all image models tested.
Shorts GeneratorlongstoriesUses LongStories AI to generate high-quality content from text prompts. Generates engaging short stories, similar to Youtube Shorts, TikTok clips etc, on any subject you want. Offers many customization options. Note: generation can take from 30 seconds to a few minutes.
Shorts Generator for Kidslongstories-kidsGenerates engaging short stories for kids on any subject you want. Offers many customization options. Note: generation can take from 30 seconds to a few minutes.
Flux Pro V1.1flux-pro/v1.1Excellent image quality, prompt adherence, and output diversity.
Imagen V3imagen-3.0-generate-002Google's highest quality text-to-image model with fine detail, rich lighting, and excellent text rendering capabilities.
Flux Pro V1.1 Ultraflux-pro/v1.1-ultra4K version of Flux Pro V1.1. Excellent image quality, prompt adherence, and output diversity.
Ideogram V2ideogram-ai/ideogram-v2An excellent image model with state of the art inpainting, prompt comprehension and especially text rendering.
Flux Loraflux-loraFLUX.1 [dev] with LoRA support, fast and high-quality image generation with the option to use LORAs for specific styles.
Flux Devflux-devSlightly faster and much cheaper than Flux Pro with similar output quality.
Flux Schnellflux/schnellFast and high-quality image generation - the cheaper version of the Flux range of models.
SD 3.5 Largestable-diffusion-v35-largeStable Diffusion's newest model. Generates a wide variety of images reflecting different styles without complex prompting.
Ideogram V2 Turboideogram-ai/ideogram-v2-turboA fast image model with state of the art inpainting, prompt comprehension and especially text rendering.
Flux Realismflux-realismIncredibly photorealistic image generation. Generate people, animals, landscapes that are hard to distinguish from reality.
DALL-E-3dall-e-3OpenAI's most well-known image model.
DALL-E-3 HDdall-e-3-hdOpenAI's most well-known image model, now in HD quality.
SD 3.5 Large Turbostable-diffusion-v35-large/turboTurbo version of Stable Diffusion's newest model. Faster and cheaper performance while still maintaining great prompt adherence and quality.
Playground V2.5playground-v25Playground V2.5 outperforms SDXL in many user tests. Suitable for a broad range of images.
Proteusproteus-v0.2A versatile image generation model with high-quality outputs.
PromptchanpromptchanThe best NSFW image generation. High-quality image generation with lots of customization options.
Flux Dev Uncensoredflux-dev-uncensoredFlux Dev Uncensored version for unrestricted image generation
Fluentlyfluently-xlFluently model for high-quality image generation
Lustify SDXLlustify-sdxlHigh-quality NSFW image generation model based on SDXL architecture
Uber RealisticuberRealisticPornMerge_urpmv12_4979.safetensorsGenerates realistic-looking NSFW images.
Stable Diffusion 3 Mediumsd3_base_medium.safetensorsExcels at photorealism, typography, and prompt following. Works best in 1024x1024.
Dreamshaper XLdreamshaper_8_93211.safetensorsDreamshaper generates realistic and anime/illustration-style images, and is best suited to sci-fi and fantasy scenes.
ReV AnimatedrevAnimated_v122.safetensorsReV Animated specialized in fantasy, anime and semi-realistic landscapes.
Stable Diffusion XLfast-sdxlCheap and powerful text-to-image model that generates pictures rapidly.
Flux Pro V1flux-proOlder version of Flux V1.1. Exceptional quality and prompt adherence.