Modern artificial intelligence technologies open new opportunities for business and science. A comprehensive analysis of major language models with a focus on their capabilities, test results, and technical parameters. Developers of advanced language models:
| Language Model | Provider | Rating | Release Date | Knowledge Cut-off Date | Open Source | API Providers | Modalities | Pricing Input | Pricing Output | MMLU | MMLU-Pro | MMMU | HellaSwag | HumanEval | MATH | GPQA | IFEval | Mobile Application | Pros | Cons |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 2.5 Pro | ⭐️ 4.6 | Mar 25, 2025 | - | No | Google AI Studio, Vertex AI, Gemini app | text images voice video | Not available | Not available | Not available | Not available | 81.7% Source | Not available | Not available | Not available | 84.0% Diamond Science Source | Not available | Google Play Apple Apps | 4870 | 204 | |
| Llama 4 Scout | Meta | ⭐️ 4.5 | Apr 05, 2025 | 2025-04 | Yes (Source) | Meta AI, Hugging Face, Fireworks, Together, DeepInfra | text images video | Not available | Not available | Not available | 74.3% Reasoning & Knowledge Source | 69.4% Image Reasoning Source | Not available | Not available | Not available | 57.2% Diamond Source | Not available | - | 4193 | 846 |
| DeepSeek-R1 | DeepSeek | ⭐️ 4.4 | Jan 21, 2025 | Unknown | Yes | DeepSeek, HuggingFace | text | $0.55 per million tokens | $2.19 per million tokens | 90.8% Pass@1 Source | 84% EM Source | - | - | - | - | 71.5% Pass@1 Source | 83.3% Prompt Strict Source | Google Play Apple Apps | 1224 | 21 |
| Grok 3 Beta | xAI | ⭐️ 4.4 | Jan 19, 2025 | 2025-01 | No | xAI | text images video | Not available | Not available | Not available | 79.9% Base model Source | 78% With Think mode Source | Not available | Not available | Not available | 84.6% With Think mode, Diamond Source | Not available | Google Play Apple Apps | 3791 | 261 |
| GPT-4.5 | OpenAI | ⭐️ 4.4 | Feb 27, 2025 | 2023-10 | No | OpenAI, Azure OpenAI Service | text images | $75.00 per million tokens | $150.00 per million tokens | Not available | Not available | 74.4% Source | Not available | Not available | Not available | 71.4% science Source | Not available | Google Play Apple Apps | 4249 | 760 |
| Qwen2.5-VL-32B | Alibaba | ⭐️ 4.4 | Mar 25, 2025 | Unknown | Yes (Source) | - | text images video | $0 | $0 | 78.4% Source | 49.5% | 70% | Not available | Not available | 82.2% | 46.0% Diamond | Not available | - | 3200 | 126 |
| o4-mini | OpenAI | ⭐️ 4.3 | Apr 16, 2025 | - | No | OpenAI API | text images | $1.10 per million tokens | $4.40 per million tokens | fort | - | 81.6% Source | - | 14.28% Source | - | 81.4% Source | - | Google Play Apple Apps | 4293 | 828 |
| Nova Pro | Amazon | ⭐️ 4.2 | Dec 02, 2024 | Purposefully not disclosed | No | Amazon Bedrock | text images video | $0.80 per million tokens | $3.20 per million tokens | 85.9% CoT Source | Not available | Not available | Not available | 89% pass@1 Source | 76.6% CoT Source | 46.9% Main Source | 92.1% Source | - | 3775 | 982 |
| Llama 4 Maverick | Meta | ⭐️ 4.2 | Apr 05, 2025 | 2024-08 | Yes (Source) | Meta AI, Hugging Face, Fireworks, Together, DeepInfra | text images video | Not available | Not available | Not available | 80.5% Source | 73.4% Source | Not available | Not available | Not available | 69.8% Diamond Source | Not available | - | 3589 | 252 |
| Claude 3.5 Haiku | Anthropic | ⭐️ 4.1 | Nov 04, 2024 | 01.04.2024 | No | Anthropic, AWS Bedrock, Vertex AI | text | $0.80 per million tokens | $4.00 | Not available | 65% 0-shot CoT Source | Not available | Not available | 88.1% 0-shot Source | 69.4% 0-shot CoT Source | Not available | Not available | Google Play Apple Apps | 3369 | 663 |
| o3-mini | OpenAI | ⭐️ 4.1 | Jan 31, 2025 | Unknown | No | OpenAI API | text | $1.10 per million tokens | $4.40 per million tokens | 86.9% pass@1, high effort Source | Not available | Not available | Not available | Not available | 97.9% pass@1, high effort Source | 79.7% 0-shot, high effort Source | Not available | Google Play Apple Apps | 4552 | 194 |
| Claude 3.7 Sonnet - Extended Thinking | Anthropic | ⭐️ 4.1 | Feb 24, 2025 | - | No | Claude.ai, Anthropic API, Amazon Bedrock, Google Cloud Vertex AI | text images | $3.00 per million tokens | $15.00 per million tokens | Not available | Not available | 75% Source | Not available | Not available | 96.2% Source | 84.8% Diamond Source | 93.2% Source | Google Play Apple Apps | 4424 | 831 |
| Gemini 2.0 Pro | ⭐️ 4 | Dec 11, 2024 | 08.2024 | No | Google AI Studio, Vertex AI | text images voice video | $0.10 per million tokens | $0.40 per million tokens | Not available | 79.1% Source | 72.7% Source | Not available | Not available | 91.8% Source | 64.7% Diamond Source | Not available | Google Play Apple Apps | 4876 | 757 | |
| Gemini 2.0 Flash | ⭐️ 4 | Dec 11, 2024 | 08.2024 | No | Google AI Studio, Vertex AI | text images voice video | $0.10 per million tokens | $0.40 per million tokens | Not available | 77.6% Source | 71.7% Source | Not available | Not available | 90.9% Source | 60.1% Diamond Source | Not available | Google Play Apple Apps | 4052 | 628 | |
| Claude 3.7 Sonnet | Anthropic | ⭐️ 4 | Feb 24, 2025 | - | No | Claude.ai, Anthropic API, Amazon Bedrock, Google Cloud Vertex AI | text images | $3.00 per million tokens | $15.00 per million tokens | Not available | Not available | 71.8% Source | Not available | Not available | 82.2% Source | 68% Diamond Source | 90.8% Source | Google Play Apple Apps | 4934 | 979 |
| Gemini 2.0 Flash Thinking | ⭐️ 3.9 | Dec 19, 2024 | 04.2024 | No | Google AI Studio, Vertex AI, Gemini API | text images | Not available | Not available | Not available | Not available | 75.4% Source | Not available | Not available | Not available | 74.2% Diamond Science Source | Not available | Google Play Apple Apps | 4466 | 870 | |
| Llama 3.3 70B Instruct | Meta | ⭐️ 3.9 | Dec 06, 2024 | 12.2024 | Yes | Fireworks, Together, DeepInfra, Hyperbolic | text | $0.23 per million tokens | $0.40 per million tokens | 86% 0-shot, CoT Source | 68.9% 5-shot, CoT Source | Not available | Not available | 88.4% pass@1 Source | 77% 0-shot, CoT Source | 50.5% 0-shot, CoT Source | 92.1% Source | - | 3691 | 541 |
| Llama 3.1 Nemotron 70B Instruct | NVIDIA | ⭐️ 3.9 | Oct 15, 2023 | - | Yes | OpenRouter | text | $0.35 per million tokens | $0.40 per million tokens | 85% 5-shot Source | Not available | Not available | Not available | 75% Source | 71% Source | Not available | Not available | - | 4028 | 265 |
| Command A | Cohere | ⭐️ 3.9 | Mar 14, 2025 | - | Yes | Cohere, Hugging Face, Major cloud providers | text | $2.50 per million tokens | $10.00 per million tokens | 85.5% Source | Not available | Not available | Not available | Not available | 80% Source | 50.8% Source | 90.9% Source | - | 3020 | 186 |
| Nova Lite | Amazon | ⭐️ 3.8 | Dec 02, 2024 | Purposefully not disclosed | No | Amazon Bedrock | text images video | $0.06 per million tokens | $0.24 per million tokens | 80.5% CoT Source | Not available | Not available | Not available | 85.4% pass@1 Source | 73.3% CoT Source | 42% Main Source | 89.7% Source | - | 3967 | 819 |
| Mistral Large 2 | Mistral AI | ⭐️ 3.8 | Jun 24, 2024 | Unknown | Yes | Azure AI, AWS Bedrock, Google AI Studio, Vertex AI, Snowflake Cortex | text | $3.00 per million tokens | $9.00 per million tokens | 84% 5-shot Source | 50.69% Source | Not available | Not available | Not available | 1.13% Source | 24.94% | 84.01% | - | 4923 | 871 |
| Nova Micro | Amazon | ⭐️ 3.6 | Dec 02, 2024 | Purposefully not disclosed | No | Amazon Bedrock | text | $0.04 per million tokens | $0.14 per million tokens | 77.6% CoT Source | - | - | - | 81.1% pass@1 Source | 69.3% CoT Source | 40% Main Source | 87.2% Source | - | 2322 | 480 |
Modern large language models (LLM) continue to evolve, providing users with powerful tools for text processing and generation. This section features key development companies, their technologies, and capabilities.
Developers of Claude – a language model focused on safety and reliability.
Creators of GPT-4, one of the most powerful language models for text generation and analysis.
Developers of Gemini – an advanced model integrated with search technologies.
Developers of Llama – an open language model for research and commercial applications.
Compare AI. Test. Benchmarks. Mobile Apps Chatbots, Sketch
Copyright © 2025 All Right Reserved.