Small (4–14B) AI models

All 189 open small models. Runs on a good laptop or consumer GPU. Compare calculated local requirements, provider-published rented-GPU rates, and dated API rates or clearly labeled estimates.

• Llama 3.1 8B Instruct — 8B, Meta · $0.07/1M API
• Qwen2.5 7B Instruct — 7B, Alibaba · $0.15/1M API
• Mistral 7B Instruct v0.3 — 7.2B, Mistral AI · $0.20/1M API last-known
• Qwen2.5-Coder 7B Instruct — 7B, Alibaba · $0.16/1M API est.
• Qwen3 4B — 4B, Qwen · $0.13/1M API est.
• Gemma 2 9B Instruct — 9B, Google · $0.06/1M API last-known
• Qwen3-8B — 8.2B, Alibaba · $0.29/1M API
• Gemma 3 12B — 12B, Google · $0.10/1M API
• Qwen3 4B Instruct 2507 — 4B, Qwen · $0.13/1M API est.
• Qwen2-VL 7B Instruct — 8B, Alibaba · $0.16/1M API est.
• Rio 3.0 Open Mini — 4B, prefeitura-rio · $0.13/1M API est.
• Hermes 3 — Llama 3.1 8B — 8B, Nous Research · $0.16/1M API est.
• Meta Llama 3 8B Instruct — 8B, meta-llama · $0.16/1M API est.
• Mistral 7B Instruct v0.2 — 7.2B, mistralai · $0.16/1M API est.
• Llama 3.1 8B — 8B, meta-llama · $0.16/1M API est.
• Meta Llama 3 8B — 8B, meta-llama · $0.16/1M API est.
• NVIDIA Nemotron 3 Nano 4B BF16 — 4B, nvidia · $0.13/1M API est.
• Llama 3.1 8B Instruct (Abliterated) — 8B, mlabonne (community) · $0.16/1M API est.
• LLaDA 8B Instruct — 8B, GSAI-ML · $0.16/1M API est.
• granite 4.1 8b — 8.8B, ibm-granite · $0.08/1M API
• Mistral 7B v0.1 — 7.2B, mistralai · $0.16/1M API est.
• Qwen3 4B Base — 4B, Qwen · $0.13/1M API est.
• Dolphin 3.0 — Llama 3.1 8B — 8B, Cognitive Computations · $0.16/1M API est.
• Llama 2 7b hf — 6.7B, meta-llama · $0.15/1M API est.
• Qwen2 7B Instruct — 7.6B, Qwen · $0.16/1M API est.
• Qwen3 4B Thinking 2507 — 4B, Qwen · $0.13/1M API est.
• Qwen2.5 7B — 7.6B, Qwen · $0.16/1M API est.
• NVIDIA Nemotron Nano 9B v2 — 8.9B, nvidia · $0.17/1M API est.
• Bielik 11B v3.0 Instruct — 11.2B, speakleash · $0.19/1M API est.
• Nemotron Labs Diffusion 8B Base — 8.5B, nvidia · $0.17/1M API est.
• deepseek coder 7b instruct v1.5 — 6.9B, deepseek-ai · $0.16/1M API est.
• DeepSeek R1 Distill Llama 8B — 8B, deepseek-ai · $0.16/1M API est.
• Qwen3 8B Base — 8.2B, Qwen · $0.17/1M API est.
• saiga llama3 8b — 8B, IlyaGusev · $0.16/1M API est.
• DeepSeek R1 Distill Qwen 7B — 7.6B, deepseek-ai · $0.16/1M API est.
• Dream v0 Instruct 7B — 7.6B, Dream-org · $0.16/1M API est.
• Qwen2.5 Coder 7B — 7.6B, Qwen · $0.16/1M API est.
• VLM2Vec Full — 4.1B, TIGER-Lab · $0.13/1M API est.
• Meta Llama 3.1 8B Instruct — 8B, unsloth · $0.16/1M API est.
• falcon 7b — 7.2B, tiiuae · $0.16/1M API est.
• DeepSeek R1 0528 Qwen3 8B — 8.2B, deepseek-ai · $0.17/1M API est.
• gemma 4 E4B it OBLITERATED — 8B, OBLITERATUS · $0.16/1M API est.
• Darwin 9B NEG — 9.7B, ansulev · $0.18/1M API est.
• Vikhr Nemo 12B Instruct R 21 09 24 — 12.2B, Vikhrmodels · $0.20/1M API est.
• Qwen3Guard Gen 4B — 4.4B, Qwen · $0.14/1M API est.
• Mistral 7B Instruct v0.1 — 7.2B, mistralai · $0.16/1M API est.
• Llama 2 7b chat hf — 6.7B, meta-llama · $0.15/1M API est.
• Phi 3 vision 128k instruct — 4.1B, microsoft · $0.13/1M API est.
• CodeLlama 7b hf — 6.7B, codellama · $0.15/1M API est.
• NVIDIA Nemotron Nano 9B v2 Japanese — 8.9B, nvidia · $0.17/1M API est.
• Gemma 4 E4B DECKARD HERETIC NVFP4 — 6.2B, AEON-7 · $0.15/1M API est.
• talkie 1930 13b it hf — 13.3B, lewtun · $0.21/1M API est.
• Qwen2.5 7B Instruct (Abliterated) — 7B, huihui-ai (community) · $0.16/1M API est.
• llama 7b — 6.7B, huggyllama · $0.15/1M API est.
• pythia 12b — 12B, EleutherAI · $0.20/1M API est.
• Fanar 1 9B Instruct — 8.8B, QCRI · $0.17/1M API est.
• Qwen3 8B — 8.2B, unsloth · $0.17/1M API est.
• gemma4 e4b claims comparison — 7.9B, k-chirkunov · $0.16/1M API est.
• Llama 3 Taiwan 8B Instruct — 8B, yentinglin · $0.16/1M API est.
• Phi mini MoE instruct — 7.6B, microsoft · $0.16/1M API est.
• Olmo 3 7B Instruct SFT — 7.3B, allenai · $0.16/1M API est.
• Apertus 70B Instruct 2509 quantized.w4a16 — 11.3B, RedHatAI · $0.19/1M API est.
• deepseek coder 6.7b instruct — 6.7B, deepseek-ai · $0.15/1M API est.
• wildguard — 7.2B, allenai · $0.16/1M API est.
• zephyr 7b beta — 7.2B, HuggingFaceH4 · $0.16/1M API est.
• MiMo 7B Base — 7.8B, XiaomiMiMo · $0.16/1M API est.
• LLaDA 1.5 — 8B, GSAI-ML · $0.16/1M API est.
• Nemotron Labs Diffusion 8B — 8.5B, nvidia · $0.17/1M API est.
• Olmo 3 7B Instruct — 7.3B, allenai · $0.16/1M API est.
• pythia 6.9b — 7B, EleutherAI · $0.16/1M API est.
• MiniCPM4.1 8B — 8.2B, openbmb · $0.17/1M API est.
• avibe — 7.9B, AvitoTech · $0.16/1M API est.
• llama3.1 typhoon2 8b instruct — 8B, typhoon-ai · $0.16/1M API est.
• xflux text encoders — 4.8B, XLabs-AI · $0.14/1M API est.
• Llama 3.1 Nemotron Nano 8B v1 — 8B, nvidia · $0.16/1M API est.
• vllm translategemma 12b it — 13.2B, chbae624 · $0.21/1M API est.
• Llama 3 8B UltraMedical — 8B, TsinghuaC3I · $0.16/1M API est.
• granite guardian 3.3 8b — 8.2B, ibm-granite · $0.17/1M API est.
• Qwen2.5 Math 7B Instruct — 7.6B, Qwen · $0.16/1M API est.
• NVIDIA Nemotron Nano 12B v2 — 12.3B, nvidia · $0.20/1M API est.
• Meta Llama 3 8B Instruct — 8B, NousResearch · $0.16/1M API est.
• Meta Llama 3.1 8B Instruct — 8B, NousResearch · $0.16/1M API est.
• EXAONE 3.5 7.8B Instruct — 7.8B, LGAI-EXAONE · $0.16/1M API est.
• Tarsier 7b — 7.1B, omni-research · $0.16/1M API est.
• Olmo 3 1025 7B — 7.3B, allenai · $0.16/1M API est.
• falcon 7b instruct — 7.2B, tiiuae · $0.16/1M API est.
• Qwen1.5 7B — 7.7B, Qwen · $0.16/1M API est.
• falcon mamba 7b — 7.3B, tiiuae · $0.16/1M API est.
• granite 4.0 h tiny — 6.9B, ibm-granite · $0.16/1M API est.
• granite 4.0 tiny preview — 6.7B, ibm-granite · $0.15/1M API est.
• Qwen3 14B NVFP4 — 8.2B, nvidia · $0.17/1M API est.
• Llama 3.1 8B Instruct — 8B, unsloth · $0.16/1M API est.
• internlm2 5 7b chat — 7.7B, internlm · $0.16/1M API est.
• Lucie 7B Instruct v1.1 — 6.7B, OpenLLM-France · $0.15/1M API est.
• Qwen3 8B NVFP4 — 4.7B, nvidia · $0.14/1M API est.
• gemma 2 9b it — 9.2B, unsloth · $0.17/1M API est.
• typhoon2.5 qwen3 4b — 4B, typhoon-ai · $0.13/1M API est.
• Llama 3.1 Swallow 8B Instruct v0.5 — 8B, tokyotech-llm · $0.16/1M API est.
• Apertus 8B Instruct 2509 — 8.1B, swiss-ai · $0.16/1M API est.
• hf moshiko — 7.8B, kmhf · $0.16/1M API est.
• granite 3.1 8b instruct — 8.2B, ibm-granite · $0.17/1M API est.
• Josiefied Qwen3 VL 4B Instruct abliterated beta v1 — 4.4B, Goekdeniz-Guelmez · $0.14/1M API est.
• Llama 2 7b hf — 6.7B, NousResearch · $0.15/1M API est.
• Llama 2 13b chat hf — 13B, meta-llama · $0.20/1M API est.
• OLMoE 1B 7B 0125 Instruct — 6.9B, allenai · $0.16/1M API est.
• OLMoE 1B 7B 0924 — 6.9B, allenai · $0.16/1M API est.
• LLaDA 8B Base — 8B, GSAI-ML · $0.16/1M API est.
• Humanish Roleplay Llama 3.1 8B — 8B, vicgalle · $0.16/1M API est.
• Qwen 7B Chat — 7.7B, Qwen · $0.16/1M API est.
• Gemma 4 12B OBLITERATED — 12B, OBLITERATUS · $0.20/1M API est.
• Gemma 4 E4B it NVFP4 — 6B, bg-digitalservices · $0.15/1M API est.
• llm jp 4 8b thinking — 8.6B, llm-jp · $0.17/1M API est.
• Qwythos 9B Claude Mythos 5 1M — 9.4B, empero-ai · $0.18/1M API est.
• GDN primed HQwen3 8B Instruct — 8.5B, amazon · $0.17/1M API est.
• MiniCPM4 8B — 8.2B, openbmb · $0.17/1M API est.
• mistralai Mistral Nemo Instruct 2407 — 12.2B, SillyTilly · $0.20/1M API est.
• gemma 4 12B agentic fable5 composer2.5 v2 3.5x tau2 — 12B, yuxinlu1 · $0.20/1M API est.
• Nemotron H 8B Base 8K — 8.1B, nvidia · $0.16/1M API est.
• Qwen2 7B — 7.6B, Qwen · $0.16/1M API est.
• Bielik 11B v2.3 Instruct — 11.2B, speakleash · $0.19/1M API est.
• Meta Llama 3 8B — 8B, NousResearch · $0.16/1M API est.
• CodeQwen1.5 7B Chat — 7.3B, Qwen · $0.16/1M API est.
• T lite it 2.1 — 8.2B, t-tech · $0.17/1M API est.
• Qwen2.5 7B Instruct 1M — 7.6B, Qwen · $0.16/1M API est.
• TimeOmni 1 7B — 7.6B, anton-hugging · $0.16/1M API est.
• Ornith 1.5 9B — 9.7B, ornith-ai · $0.18/1M API est.
• Falcon3-10B Instruct — 10B, TII · $0.18/1M API est.
• MiMo 7B RL — 7.8B, XiaomiMiMo · $0.16/1M API est.
• SeeClick — 9.7B, cckevinn · $0.18/1M API est.
• LFM2.5 8B A1B — 8.5B, LiquidAI · $0.17/1M API est.
• Qwen3Guard Gen 8B — 8.2B, Qwen · $0.17/1M API est.
• gemma 2 9b — 9.2B, google · $0.17/1M API est.
• HarmBench Llama 2 13b cls — 13B, cais · $0.20/1M API est.
• Qwen2.5 7B Instruct — 7.6B, unsloth · $0.16/1M API est.
• Meta Llama 3.1 8B Instruct quantized.w4a16 — 8B, RedHatAI · $0.16/1M API est.
• Qwen2.5 Math 7B — 7.6B, Qwen · $0.16/1M API est.
• Agents A1 4B — 4.5B, InternScience · $0.14/1M API est.
• Llama Guard 3 8B — 8B, meta-llama · $0.16/1M API est.
• internlm3 8b instruct — 8.8B, internlm · $0.17/1M API est.
• granite 3.0 8b instruct — 8.2B, ibm-granite · $0.17/1M API est.
• granite 3.3 8b instruct — 8.2B, ibm-granite · $0.17/1M API est.
• SOLAR 10.7B Instruct v1.0 — 10.7B, upstage · $0.19/1M API est.
• Qwen Marketing — 8.2B, marketeam · $0.17/1M API est.
• internlm2 chat 7b — 7.7B, internlm · $0.16/1M API est.
• OLMo 2 1124 7B Instruct — 7.3B, allenai · $0.16/1M API est.
• llava onevision qwen2 7b ov — 8B, lmms-lab · $0.16/1M API est.
• granite guardian 4.1 8b — 8.4B, ibm-granite · $0.17/1M API est.
• LFM2 8B A1B — 8.3B, LiquidAI · $0.17/1M API est.
• Olmo 3 7B Think — 7.3B, allenai · $0.16/1M API est.
• GLM 4 9B 0414 — 9.4B, zai-org · $0.18/1M API est.
• Qwen3.6 27B MTPLX Optimized Speed — 4.7B, Youssofal · $0.14/1M API est.
• gemma 4 12B coder fable5 composer2.5 v1 — 12B, yuxinlu1 · $0.20/1M API est.
• Olmo Hybrid 7B — 7.4B, allenai · $0.16/1M API est.
• Foundation Sec 8B Reasoning — 8B, fdtn-ai · $0.16/1M API est.
• Huihui gemma 4 12B agentic fable5 abliterated — 12B, huihui-ai · $0.20/1M API est.
• gemma 7b — 8.5B, google · $0.17/1M API est.
• MN 12B Mag Mell R1 — 12.2B, inflatebot · $0.20/1M API est.
• Mellum2 12B A2.5B Thinking — 12.1B, JetBrains · $0.20/1M API est.
• c4ai command r7b 12 2024 — 8B, CohereLabs · $0.16/1M API est.
• Qwable 9B Claude Fable 5 — 9.4B, empero-ai · $0.18/1M API est.
• mythos 9b unhinged — 8.2B, King3Djbl · $0.17/1M API est.
• VulnLLM R 7B — 7.6B, Virtue-AI-HUB · $0.16/1M API est.
• rnj 1 — 8.3B, EssentialAI · $0.17/1M API est.
• OpenCodeInterpreter DS 6.7B — 6.7B, m-a-p · $0.15/1M API est.
• aya expanse 8b — 8B, CohereLabs · $0.16/1M API est.
• Nanbeige4.2 3B — 4.2B, Nanbeige · $0.13/1M API est.
• qwen3 4b heretic — 4B, DreamFast · $0.13/1M API est.
• Qwen3.8 27B MTPLX Optimized Quality — 8B, Youssofal · $0.16/1M API est.
• Llama 2 7b chat hf — 6.7B, NousResearch · $0.15/1M API est.
• Llama 3.1 Nemotron Safety Guard 8B v3 — 8B, nvidia · $0.16/1M API est.
• Mellum2 12B A2.5B Base — 12.1B, JetBrains · $0.20/1M API est.
• starcoder2 7b — 7.2B, bigcode · $0.16/1M API est.
• AFM 4.5B — 4.6B, arcee-ai · $0.14/1M API est.
• Qwythos 9B v2 — 9.7B, empero-ai · $0.18/1M API est.
• Saul 7B Instruct v1 — 7.2B, Equall · $0.16/1M API est.
• Qwen3.8 27B MTPLX Optimized Speed — 6.1B, Youssofal · $0.15/1M API est.
• Ling 3.0 tiny — 7.9B, inclusionAI · $0.16/1M API est.
• Apertus 8B 2509 — 8.1B, swiss-ai · $0.16/1M API est.
• gemma 3 12b it heretic — 12.2B, DreamFast · $0.20/1M API est.
• DeepHat V1 7B — 7.6B, DeepHat · $0.16/1M API est.
• Nemotron Labs Diffusion 14B — 13.5B, nvidia · $0.21/1M API est.
• Qwen3.8 9B Distill — 9.7B, empero-ai · $0.18/1M API est.
• Qwen3.8 27B Abliterated MTPLX Optimized Speed — 6.1B, PocketAiHub · $0.15/1M API est.
• NeuralDaredevil 8B abliterated — 8B, mlabonne · $0.16/1M API est.
• L3 8B Stheno v3.2 — 8B, Sao10K · $0.16/1M API est.
• OpenCoder 8B Instruct — 7.8B, infly · $0.16/1M API est.
• Llama 3.1 8B Lexi Uncensored V2 — 8B, Orenguteng · $0.16/1M API est.
• gemma 3 12b it heretic v2 — 12.2B, DreamFast · $0.20/1M API est.
• ALLaM 7B Instruct preview — 7B, humain-ai · $0.16/1M API est.

Other sizes

Flagship (80B+) · Large (34–80B) · Medium (14–34B) · Tiny (under 4B) · All models

Compare → · LLM cost calculator →

Open the free LLM cost calculator → · Rate table dated 2026-08-24. $0 markup, your own accounts, we never resell compute. © 2026 Cynosure LLC.