Tools
Cost Calculator
What a month costs, by model. Every figure here is measured — nothing is estimated. See Methodology for exactly how it is calculated.
How much will you use it?
A task is one Artificial Analysis Intelligence Index task — one self-contained question or job. The anchors are rough guides, not measurements; type an exact figure if you know it.
Monthly cost
- GPT-5.6 Luna$84.75$0.0113/task (medium effort)
- GPT-6 Luna$150$0.02/task (medium effort)
- DeepSeek-V3$173$0.0231/task
- Llama 4 Maverick$215$0.0286/task
- Gemma 4$257$0.0343/task
- Hy3$268$0.0357/task
- DeepSeek-V4$336$0.0448/task (max effort)
- Mistral Large 3$449$0.0599/task
Figures measured at different reasoning efforts are not strictly comparable — see Methodology.
61 of 120 active models have no measured cost-per-task figure and are excluded (14 more are deprecated and not shown)
Muse Wonder Full, Kolibri-1, Ling-3.1-flash, MiMo-V2.6-Flash, Qwen3.8-Omni-Flash, Laya, Jev, Fugu Max, Fugu Ultra v2.0, Atria Dawn Preview, Kimi K2.8 Preview, HelixWorld, Gemini 3.8 Flash Cyber, Qwen3.8-Max-0902, Hy4 preview, Ornith-1.5-397B, Qwen3.8-27B, GLM-5.3, Qwen3.8-2.4T-A95B, DeepSeek-V4-Pro-0813, Motif-3, Muse Glimmer, GPT-5.6-Cyber, Mach-1 Additive 35B, Qwen3.7 Flash, Laguna S 2.1, Motif-3-Beta, Qwen3.8 Max, Inkling, Soofi S 30B-A3B, Laguna XS 2.1, Seed 2.1 Pro, MiniMax-M3, Command A+, ERNIE 5.1, Ling-2.6-1T, Muse Spark, Seed 2.0 Pro, MiniMax-M2.5, ERNIE 5.0, Nova 2 Pro, Claude Opus 4.5, Gemini 3 Pro, Grok 4.1, Kimi K2 Thinking, MiniMax-M2, DeepSeek-V3.2, Qwen3-Max, gpt-oss-120b, gpt-oss-20b, Kimi K2, Grok 4, Mistral Medium 3, Nova Premier, Qwen3-235B-A22B, GPT-4.1, Command A, Grok 3, Gemini 2.0 Flash, o1, GPT-4o