liminfoLLM Model Comparison
Free web tool: LLM Model Comparison
| Model | Provider | Params | Context | Input/1M | Output/1M | MMLU | Strengths |
|---|---|---|---|---|---|---|---|
| GPT-4.1 | OpenAI | ~1.8T | 1M | $2.00 | $8.00 | 90.2 | Multimodal, Coding, Instruction following |
| GPT-4o | OpenAI | ~1.8T | 128K | $2.50 | $10.00 | 88.7 | Multimodal, Coding, Reasoning |
| GPT-4o mini | OpenAI | - | 128K | $0.15 | $0.60 | 82 | Cost-effective, Fast |
| Claude Opus 4 | Anthropic | - | 200K | $15.00 | $75.00 | 91.3 | Top reasoning, Coding, Analysis |
| Claude Sonnet 4.5 | Anthropic | - | 200K | $3.00 | $15.00 | 90 | Coding, Analysis, Balanced performance |
| Claude 3.5 Haiku | Anthropic | - | 200K | $0.80 | $4.00 | 84 | Fast response, Cost-effective |
| Gemini 2.5 Pro | - | 1M | $1.25 | $10.00 | 90.8 | Reasoning, Coding, Large context | |
| Gemini 2.0 Flash | - | 1M | $0.10 | $0.40 | 85.2 | Ultra-large context, Fast | |
| Llama 4 | Meta | 405B MoE | 256K | Open | Open | 89 | Open source, Multimodal, MoE |
| Llama 3.1 405B | Meta | 405B | 128K | Open | Open | 87.3 | Open source, Self-hosting |
| DeepSeek V3 | DeepSeek | 671B MoE | 128K | $0.27 | $1.10 | 88.5 | MoE, Coding, Cost-effective |
| Mistral Large | Mistral | 123B | 128K | $2.00 | $6.00 | 86 | Multilingual, Coding |
| Command R+ | Cohere | 104B | 128K | $2.50 | $10.00 | 83 | RAG, Search augmentation |
| Qwen 2.5 72B | Alibaba | 72B | 128K | Open | Open | 85.8 | Open source, Multilingual |
Cost Calculator
About LLM Model Comparison
The LLM Model Comparison tool lets you instantly compare 11 of the most widely used large language models in a single sortable table. The models covered include GPT-4o and GPT-4o mini from OpenAI, Claude 3.5 Sonnet and Claude 3.5 Haiku from Anthropic, Gemini 2.0 Flash and Gemini 1.5 Pro from Google, Llama 3.1 405B from Meta, Mistral Large, Command R+ from Cohere, DeepSeek V3, and Qwen 2.5 72B from Alibaba. For each model you can see the provider, parameter count, context window size, input price per million tokens, output price per million tokens, and a summary of its strengths.
This tool is built for AI engineers, product managers, startup founders, and researchers who need to quickly evaluate which model best fits their use case and budget. Selecting the right LLM involves balancing cost, latency, context length, and task-specific performance. For example, Gemini 1.5 Pro offers a 2 million token context window ideal for long document analysis, while GPT-4o mini and Gemini 2.0 Flash are cost-optimized for high-volume workloads. Open-source models like Llama 3.1 405B and Qwen 2.5 72B have no API cost and can be self-hosted for maximum data privacy.
The comparison data is embedded directly in the component as a static dataset and filtered entirely in the browser using JavaScript. You can search by model name or strength keywords and filter by provider (OpenAI, Anthropic, Google, Meta, Mistral, Cohere, DeepSeek, Alibaba) using the toggle buttons. The result count updates in real time as you type. No network request is made, and the page renders instantly regardless of connection speed. Pricing reflects per-million-token API rates which are useful for estimating costs before committing to a provider.
Key Features
- Covers 11 major LLM models from 8 providers: OpenAI, Anthropic, Google, Meta, Mistral, Cohere, DeepSeek, and Alibaba
- Displays context window sizes from 128K up to 2M tokens for each model
- Shows input and output pricing per million tokens for direct cost comparison
- Real-time search by model name or capability keywords (e.g., "coding", "RAG", "multilingual")
- One-click provider filter buttons to narrow down by vendor
- Open-source models clearly marked with "Open" pricing for self-hosting scenarios
- Compact table layout with horizontal scroll for comfortable use on small screens
- Strengths column in both Korean and English based on your locale setting
Frequently Asked Questions
Which LLM models are included in the comparison?
The tool covers GPT-4o, GPT-4o mini, Claude 3.5 Sonnet, Claude 3.5 Haiku, Gemini 2.0 Flash, Gemini 1.5 Pro, Llama 3.1 405B, Mistral Large, Command R+, DeepSeek V3, and Qwen 2.5 72B — 11 models from 8 leading AI providers.
What does "context window" mean and why does it matter?
The context window is the maximum number of tokens (roughly 0.75 words per token) that a model can process in a single request, including both the input prompt and the generated response. A larger context window allows you to pass in longer documents, longer conversation histories, or larger codebases. For example, Gemini 1.5 Pro supports up to 2 million tokens, which can accommodate entire books or large repositories.
How is token pricing calculated?
Pricing is expressed as cost per million tokens. Input tokens are the text you send to the model (your prompt, documents, context), while output tokens are the text the model generates. Output tokens are typically 3–6 times more expensive than input tokens. For example, GPT-4o charges $2.50 per million input tokens and $10.00 per million output tokens.
Which model is best for coding tasks?
Claude 3.5 Sonnet, GPT-4o, and DeepSeek V3 are generally considered top performers for coding. Claude 3.5 Sonnet excels at code generation, debugging, and code review with its large 200K context window. DeepSeek V3 uses a Mixture-of-Experts architecture at a much lower cost, making it attractive for high-volume coding applications.
What are open-source models and how do I use them?
Models listed with "Open" pricing — Llama 3.1 405B and Qwen 2.5 72B — are open-weight models whose model weights are freely downloadable. You can run them on your own hardware or cloud infrastructure, paying only for compute rather than per-token API fees. This is ideal for applications requiring data privacy or very high inference volumes. You can deploy them using frameworks like vLLM, Ollama, or Hugging Face TGI.
What is Mixture-of-Experts (MoE) architecture?
MoE models like DeepSeek V3 (671B MoE) route each token through only a subset of their total parameters during inference. This means the model achieves performance comparable to dense models with many more active parameters, but at significantly lower inference cost and latency. The "671B" refers to total parameters, while only a fraction are activated per token.
Which model is best for processing very long documents?
For extremely long documents, Gemini 1.5 Pro with its 2M token context window is the clear leader. Gemini 2.0 Flash offers 1M tokens at a very low price. Claude 3.5 Sonnet and Haiku both support 200K tokens, which is sufficient for most long-form use cases. All OpenAI GPT-4 models support 128K tokens.
Is the pricing data in this tool up to date?
The pricing data reflects API rates at the time the tool was built. LLM pricing changes frequently as providers compete and optimize their infrastructure. Always verify current pricing on the official provider dashboards before making production budget decisions. This table is best used as a quick reference for ballpark cost estimation and provider comparison.