/
 Hy-MT2 Plus (Tencent Cloud)TencentJSON | $0.295/1Mvia DevPass (LLM Gateway) | 8K | $0.074/1M | $0.295/1M | — | Jun 12 | |
| Hy-MT2-1.8BTencentApache 2.0 | $0.177/1Mvia OpenRouter | 8K | $0.044/1M | $0.177/1M | — | May 11 | |
| Hy-MT2-30B-A3BTencentApache 2.0JSON | $0.295/1Mvia OpenRouter | 8K | $0.074/1M | $0.295/1M | — | May 11 | |
| Hy-MT2-7BTencentApache 2.0JSON | $0.295/1Mvia OpenRouter | 8K | $0.074/1M | $0.295/1M | — | May 11 | |
| Hy3TencentApache 2.0Prompt caching+3 | $0.33/1Mvia OpenRouter | 262K | $0.083/1M | $0.33/1M | — | Jul 2 | |
| Hy3 PreviewTencentOpen weightsPrompt caching+2 | $0.60/1Mvia OpenRouter | 262K | $0.18/1M | $0.60/1M | — | Apr 13 | |
| Hy4 previewTencentApache 2.0Prompt caching+3 | $2.25/1Mvia OpenRouter | 1.05M | $0.751/1M | $2.25/1M | — | Aug 27 | |
 Inkling (256K)Thinking MachinesPrompt cachingReasoning+2 | $1.50/1Mvia Venice AI | 256K | $0.563/1M | $1.50/1M | — | Jul 15 | |
| Inkling SmallThinking MachinesPrompt cachingReasoning+2 | $1.20/1Mvia Deep Infra | 1.05M | $0.45/1M | $1.20/1M | — | Jul 30 | |
 Ising Calibration 1.5 31BNVIDIAOpen weights · permissive | Free/1Mvia Nvidia | — | Free/1M | Free/1M | — | Jul 13 | |
| Jamba 1.5 LargeAI21 Labs | $8/1Mvia Amazon Bedrock | — | $2/1M | $8/1M | — | — | |
| Hy Image3.5 previewTencent | $0.019/imgvia APIMart | — | — | $0.019/img | — | — | |
 I1 FastHiDream | $0.01/MPvia Fal | — | — | $0.01/MP | — | — | |
| IdeogramIdeogram | $0.03/imgvia Fal | — | — | $0.03/img | — | — | |
| Ideogram TransparentIdeogram | $0.03/imgvia Fal | — | — | $0.03/img | — | — | |
| Ideogram V2 RemixIdeogram | $0.08/imgvia Fal | — | — | $0.08/img | — | — | |
| Ideogram V2 TurboIdeogram | $0.05/imgvia Fal | — | — | $0.05/img | — | — | |
| Ideogram V2 Turbo RemixIdeogram | $0.05/imgvia Fal | — | — | $0.05/img | — | — | |
| Ideogram V2A TurboIdeogram | $0.025/imgvia Fal | — | — | $0.025/img | — | — | |
| Ideogram V3 BalancedIdeogram | $0.06/runvia WaveSpeedAI | — | — | $0.06/run | — | — | |
| Ideogram V3 QualityIdeogram | $0.09/runvia WaveSpeedAI | — | — | $0.09/run | — | — | |
| Ideogram V3 TurboIdeogram | $0.03/runvia WaveSpeedAI | — | — | $0.03/run | — | — | |
| Ideogram V4 BalancedIdeogram | $0.06/imgvia Replicate | — | — | $0.06/img | — | — | |
| Ideogram V4 QualityIdeogram | $0.10/imgvia Replicate | — | — | $0.10/img | — | — | |
| Ideogram VectorIdeogram | $0.11/runvia WaveSpeedAI | — | — | $0.11/run | — | — | |
 Image 01MiniMax | $0.004/runvia WaveSpeedAI | — | — | $0.004/run | — | — | |
 Image 3.2Bria | $0.04/imgvia Replicate | — | — | $0.04/img | — | — | |
| Image Subject ReferenceMiniMax | $0.01/imgvia Fal | — | — | $0.01/img | — | — | |
 Imagen 4.0Google | $0.04/imgvia APIMart | — | — | $0.04/img | — | — | |
| Imagen-3-FastGoogle | $0.025/imgvia Replicate | — | — | $0.025/img | — | — | |
| Imagen-4-FastGoogle | $0.02/imgvia Replicate | — | — | $0.02/img | — | — | |
| Imagen-4-UltraGoogle | $0.06/imgvia Replicate | — | — | $0.06/img | — | — | |
 Jib Mix Qwen ImageAlibaba | $0.02/runvia WaveSpeedAI | — | — | $0.02/run | — | — | |
 Juggernaut Flux Baserundiffusion-fal | $0.035/MPvia Fal | — | — | $0.035/MP | — | — | |
| Juggernaut Flux Lightningrundiffusion-fal | $0.006/MPvia Fal | — | — | $0.006/MP | — | — | |
| Juggernaut Flux Prorundiffusion-fal | $0.055/MPvia Fal | — | — | $0.055/MP | — | — | |
| Hunyuan VideoTencentOpen weights · with terms | $0.08/runvia WaveSpeedAI | — | — | $0.08/run | — | Dec 1, 2024 | |
| Hunyuan Video InferenceTencent | $0.40/videovia Fal | — | — | $0.40/video | — | — | |
 Image To Video 2.0Vidu (Shengshu) | $0.30/runvia WaveSpeedAI | — | — | $0.30/run | — | — | |
| Image To Video Q1Vidu (Shengshu) | $0.40/runvia WaveSpeedAI | — | — | $0.40/run | — | — | |
| Image To Video Q2 ProVidu (Shengshu) | $0.055/runvia WaveSpeedAI | — | — | $0.055/run | — | — | |
| Image To Video Q2 TurboVidu (Shengshu) | $0.04/runvia WaveSpeedAI | — | — | $0.04/run | — | — | |
| InfinitalkMeituan | $0.015/secvia Kie.ai | — | — | $0.015/sec | — | — | |
 Infinity StarByteDance | $0.07/videovia Fal | — | — | $0.07/video | — | — | |
| Inkling 256KThinking Machines | $9.36/1Mvia Impossibl | — | $3.74/1M | $9.36/1M | — | — | |
| Inkling Small 256KThinking Machines | $2.89/1Mvia Impossibl | — | $1.16/1M | $2.89/1M | — | — | |
 Kandinsky5sber | $0.05/videovia Fal | — | — | $0.05/video | — | — | |
 Inworld 1.5 MaxInworld | $0.07/runvia WaveSpeedAI | — | — | $0.07/run | — | — | |
| Inworld 1.5 MiniInworld | $0.03/runvia WaveSpeedAI | — | — | $0.03/run | — | — | |
 Hyper3dHyper3D | $0.40/runvia Fal | — | — | $0.40/run | — | — |
Best AI models by category
The top three in each category by LMArena blind-vote score, with the lowest API price. Updated daily.
AI model comparison: quality vs price
Every ranked LLM plotted by its LMArena score against its lowest price per 1M tokens. Models on the line give the most quality for the money.
Scores: LMArena Text leaderboard (CC BY 4.0). Prices: AIModelFind.
- Gemma 2 9B It1207$0.03
- Gemma 3 4B IT1291$0.05
- Gemma 3 12B IT1334$0.10
- GPT OSS 120B1366$0.14
- Qwen3.5 Flash1397$0.26
- Qwen3.7 Plus1455$0.30
- Gemini 3 Flash1466$2.40
ERNIE 5.11468$2.54- Gemini Pro Latest1480$6.90
Scores: LMArena Text leaderboard (CC BY 4.0). Prices: AIModelFind.
Browse AI models by type and use
Ready-made lists of free, open-source, Chinese and new models, and models for each task.