AI Models
Track releases, compare pricing, and benchmark scores across all major providers.
Machine-readable JSON/api/modelsThe AI model landscape changes fast. New releases ship weekly, pricing drops without warning, and context windows that seemed impossible a year ago are now standard. This page tracks every major model from Anthropic, OpenAI, Google, Meta, Mistral, Cohere, and DeepSeek with current pricing (per million tokens), context window sizes, release dates, and capability tags. Click any model name to see its full detail page with benchmark scores, strengths, and use case recommendations.
Pricing data is sourced directly from provider APIs and official documentation, refreshed daily by our worker pipeline and rehydrated in your browser on each visit. For deeper analysis, see the full pricing guide, use the cost calculator for real-world cost estimates, or compare models side by side.
Everything above is priced per token because someone else runs it. If you would rather own the inference, the open-weights side of the market is tracked separately: open-weights deployment specs gives you the VRAM, GPU class, and license for each downloadable model, and best open source LLMs compares them on capability. Frontier closed models and frontier open models are both real options now, and the cost curves cross at different points depending on your volume.
Latest Releases
Claude Opus 5
Anthropic · Jul 2026
Context: 1M
Muse Spark 1.1
Meta · Jul 2026
Context: 1M
Grok 4.5
xAI · Jul 2026
Context: 500K
Kimi K3
Moonshot AI · Jul 2026
Context: 1M
Claude Fable 5
Anthropic · Jun 2026
Context: 1M
Claude Sonnet 5
Anthropic · Jun 2026
Context: 1M
GPT-5.6 Sol
OpenAI · Jun 2026
Context: 1M
GPT-5.6 Terra
OpenAI · Jun 2026
Context: 1M
Pricing Comparison
Prices in USD per 1M tokens. Last updated 2026-07-24.
TFII is the TensorFeed Intelligence Index (0 to 100), a composite over public benchmarks. See the methodology.
| Provider | Model | TFII | Input / 1M | Output / 1M | Context |
|---|---|---|---|---|---|
| Anthropic | Claude Opus 5 | n/a | $5.00 | $25.00 | 1M |
| Anthropic | Claude Fable 5 | n/a | $10.00 | $50.00 | 1M |
| Anthropic | Claude Sonnet 5 | n/a | $2.00 | $10.00 | 1M |
| Anthropic | Claude Opus 4.8 | n/a | $5.00 | $25.00 | 1M |
| Anthropic | Claude Opus 4.7 | n/a | $15.00 | $75.00 | 1M |
| Anthropic | Claude Opus 4.6 | n/a | $15.00 | $75.00 | 200K |
| Anthropic | Claude Sonnet 4.6 | n/a | $3.00 | $15.00 | 200K |
| Anthropic | Claude Haiku 4.5 | n/a | $0.80 | $4.00 | 200K |
| OpenAI | GPT-5.6 Sol | n/a | $5.00 | $30.00 | 1M |
| OpenAI | GPT-5.6 Terra | n/a | $2.50 | $15.00 | 1M |
| OpenAI | GPT-5.6 Luna | n/a | $1.00 | $6.00 | 1M |
| OpenAI | GPT-5.5 | n/a | $5.00 | $30.00 | 1M |
| OpenAI | GPT-4o | n/a | $2.50 | $10.00 | 128K |
| OpenAI | GPT-4o-mini | n/a | $0.15 | $0.60 | 128K |
| OpenAI | o1 | n/a | $15.00 | $60.00 | 200K |
| OpenAI | o3-mini | n/a | $1.10 | $4.40 | 200K |
| Gemini 2.5 Pro | n/a | $1.25 | $10.00 | 1M | |
| Gemini 2.0 Flash | n/a | $0.10 | $0.40 | 1M | |
| Gemini 3.1 Flash-Lite | n/a | $0.25 | $1.50 | 1.0M | |
| Gemini 3.5 Flash | n/a | $1.50 | $9.00 | 1.0M | |
| Meta | Muse Spark 1.1 | n/a | $1.25 | $4.25 | 1M |
| Meta | Llama 4 Scout | n/a | Free | Free | 10M |
| Meta | Llama 4 Maverick | n/a | Free | Free | 1M |
| Mistral | Mistral Medium 3.5 | n/a | $1.50 | $7.50 | 256K |
| Mistral | Mistral Large | n/a | $2.00 | $6.00 | 128K |
| Mistral | Mistral Small | n/a | $0.10 | $0.30 | 128K |
| Cohere | Command R+ | n/a | $2.50 | $10.00 | 128K |
| Cohere | Command R | n/a | $0.15 | $0.60 | 128K |
| DeepSeek | DeepSeek V4 Pro | n/a | $0.43 | $0.87 | 1M |
| DeepSeek | DeepSeek V4 Flash | n/a | $0.14 | $0.28 | 1M |
| Alibaba | Qwen3.7-Max | n/a | $2.50 | $7.50 | 1M |
| xAI | Grok 4.5 | n/a | $2.00 | $6.00 | 500K |
| xAI | Grok 4.3 | n/a | $1.25 | $2.50 | 1M |
| NVIDIA | Nemotron 3 Nano Omni | n/a | Free | Free | 256K |
| Microsoft | MAI-Code-1-Flash | n/a | $0.75 | $4.50 | 256K |
| MiniMax | MiniMax M3 | n/a | $0.30 | $1.20 | 1.0M |
| Meituan | LongCat-2.0 | n/a | Free | Free | 1M |
| Z.ai (Zhipu AI) | GLM-5.2 | n/a | $1.40 | $4.40 | 1M |
| Moonshot AI | Kimi K3 | n/a | $3.00 | $15.00 | 1M |
Open source models are free to download and self-host. Hosted API pricing varies by provider. Prices are subject to change. Check provider websites for the most current pricing.
Run It Yourself
The pricing above is what it costs to rent a model. These are what it costs to own one.
Open-Weights Deployment
VRAM per quantization, recommended GPU class, and license for every major open-weights model. Nemotron 3, Kimi K3, GLM-5.2, LongCat-2.0, DeepSeek V4, Llama 4. What it costs to run it yourself.
Best Open Source LLMs
The comparison guide: parameters, architecture, context, licensing, and what each open model is actually good for. Start here if you are choosing rather than deploying.
Inference Providers
Hosted pricing for open-weights models across third-party providers. The middle option when you want open weights without owning the GPUs.
Leaderboard Links
LMArena
Community-driven blind model comparison with Elo ratings from real user votes.
HuggingFace Open LLM Leaderboard
Automated benchmark suite for open-weight models across reasoning, math, and knowledge tasks.
MMLU Benchmark Results
Massive Multitask Language Understanding scores across 57 academic subjects.
