Find the Cheapest & Best-Fit LLM API
Real-time tracking of 58 providers, 412 models' pricing, capabilities & context window. Covering international and Chinese models, auto-updated daily at 04:00.
This site tracks public LLM API pricing and benchmarks with daily updates. Background, sources, and caveats live on the About page →
Best Value Models
Filters to models with AA Intelligence Index 40 or higher, then ranks by intelligence per dollar (AA score ÷ input price)
| # | Model | Provider | Input | Benchmark |
|---|---|---|---|---|
| 1 | MiniMax: MiniMax M3 Multimodal | | $0.300 | AA Index 44 |
| 2 | DeepSeek: DeepSeek V4 Pro | | $0.435 | AA Index 44 |
| 3 | Xiaomi: MiMo-V2.5-Pro | | $0.435 | AA Index 42 |
| 4 | MoonshotAI: Kimi K2.7 Code Multimodal | | $0.740 | AA Index 42 |
| 5 | Z.ai: GLM 5.2 | | $0.930 | AA Index 51 |
Strongest Models
Ranked by Artificial Analysis Intelligence Index, a cross-domain intelligence metric
| # | Model | Provider | Input | AA Index |
|---|---|---|---|---|
| 1 | Anthropic: Claude Fable 5 | | $10.00 | 60 |
| 2 | Anthropic: Claude Opus 4.8 | | $5.00 | 56 |
| 3 | OpenAI: GPT-5.5 | | $5.00 | 55 |
| 4 | Anthropic: Claude Opus 4.7 | | $5.00 | 54 |
| 5 | Anthropic: Claude Sonnet 5 | | $2.00 | 53 |
Cheapest Input Price
Cost in USD per million input tokens (excluding free models)
| # | Model | Provider | Input | Output |
|---|---|---|---|---|
| 1 | inclusionAI: Ling-2.6-flash | | $0.010 | $0.030 |
| 2 | IBM: Granite 4.0 Micro | | $0.017 | $0.112 |
| 3 | Mistral: Mistral Nemo | | $0.020 | $0.030 |
| 4 | Meta: Llama 3.1 8B Instruct | | $0.020 | $0.030 |
| 5 | Meta: Llama 3.2 1B Instruct | | $0.027 | $0.201 |
Fastest Output
Ranked by Artificial Analysis measured output speed (tokens/sec)
| # | Model | Provider | t/s |
|---|---|---|---|
| 1 | Inception: Mercury 2 | | 863 |
| 2 | LiquidAI: LFM2.5-1.2B-Instruct (free) | | 452 |
| 3 | StepFun: Step 3.7 Flash | | 382 |
| 4 | Amazon: Nova Micro 1.0 | | 320 |
| 5 | Google: Gemini 3.1 Flash Lite Preview | | 307 |
Free Models (42)
Models with $0 input price. Some may still charge for output — click to view full pricing details.
| Model | Provider | Output $/M | Context |
|---|---|---|---|
| Arcee AI: Trinity Large Thinking (free) | | Free | 262K |
| Baidu Qianfan: CoBuddy (free) | | Free | 131K |
| Baidu: Qianfan-OCR-Fast (free) | | Free | 66K |
| Cohere: North Mini Code (free) | | Free | 256K |
| DeepSeek: DeepSeek V4 Flash (free) | | Free | 1.05M |
Top Providers
🇺🇸 OpenAI
- Models
- 67
- Cheapest Input
- $0.050 /M tokens
🇨🇳 Qwen (Alibaba)
- Models
- 53
- Cheapest Input
- $0.033 /M tokens
🇺🇸 Google
- Models
- 38
- Cheapest Input
- $0.060 /M tokens
🇫🇷 Mistral AI
- Models
- 25
- Cheapest Input
- $0.020 /M tokens
🇺🇸 Anthropic
- Models
- 23
- Cheapest Input
- $0.250 /M tokens
🇺🇸 NVIDIA
- Models
- 14
- Cheapest Input
- $0.040 /M tokens
🇨🇳 DeepSeek
- Models
- 14
- Cheapest Input
- $0.210 /M tokens
🇨🇳 Z.ai (Zhipu)
- Models
- 14
- Cheapest Input
- $0.060 /M tokens
Last update: 2026-07-01 · Prices in USD, for reference only. Please verify with official sources.