AISOTA· State of the Art

Daily AI Decision Intelligence

Multi-source rankings blended into one score. See which model is truly State of the Art — today.

Data as of 2026-09-03Updated Sep 3, 11:02 AM

Overall Intelligence

Multi-source blend of general intelligence benchmarks into a single AISOTA score.

#	Model	Artificial Analysis	LiveBench	AISOTA Score
1SOTA	Claude Fable 5.1 (max with fallback)2 sources	94.7	98.1	96.4
22	gpt-5.5-xhigh	—	90.6	90.6
33	Claude Fable 5 (with fallback)2 sources	84.2	96.2	90.2
4	Claude Opus 5 (max)2 sources	89.5	88.7	89.1
5	gemini-3.7-flash-high	—	86.8	86.8
6	GPT-5.6 Sol (max)2 sources	78.9	92.5	85.7
7	smaug-agentic	—	84.9	84.9
8	qwen3.8-max	—	83.0	83.0
9	Muse Spark 1.3 (xhigh)2 sources	68.4	94.3	81.3
10	muse-spark-1.2-xhigh	—	77.4	77.4

Coding Ability

Multi-source blend of coding benchmarks into a single AISOTA score.

Sources:Artificial Analysis Agentic LiveBench Coding

#	Model	Artificial Analysis Agentic	LiveBench Coding	AISOTA Score
1SOTA	Claude Fable 5.1 (max with fallback)2 sources	94.7	98.1	96.4
22	smaug-agentic	—	96.2	96.2
33	Claude Opus 5 (max)2 sources	89.5	92.5	91.0
4	GLM-5.3 (max)2 sources	84.2	86.8	85.5
5	qwen3.8-max	—	84.9	84.9
6	claude-sonnet-5-xhigh-effort	—	83.0	83.0
7	gemini-3.7-flash-high	—	77.4	77.4
8	Claude Fable 5 (with fallback)2 sources	57.9	94.3	76.1
9	deepseek-v4-flash-vision-exp	—	75.5	75.5
10	GPT-5.6 Sol (max)2 sources	68.4	81.1	74.8

Value for Money

Intelligence per dollar, based on median input + output price.

Sources:AA Intelligence OpenRouter Price

#	Model	AA Intelligence	OpenRouter Price	AISOTA Score
1SOTA	GLM-5.3-Flash2 sources	42.1	75.9	59.0
22	GPT-5.6 Luna (max)2 sources	26.3	74.6	50.5
33	GPT-5.6 Sol (max)2 sources	78.9	21.7	50.3
4	Claude Opus 5 (max)2 sources	89.5	10.4	50.0
5	Claude Fable 5.1 (max with fallback)2 sources	94.7	5.0	49.9
6	Gemini 3.8 Flash (high)2 sources	52.6	46.8	49.7
7	Muse Spark 1.3 (xhigh)2 sources	68.4	24.4	46.4
8	Grok 4.6 (high)2 sources	73.7	17.7	45.7
9	Claude Fable 5 (with fallback)2 sources	84.2	4.7	44.5
10	GLM-5.3 (max)2 sources	57.9	22.7	40.3

Model Usage

API request volume per model over the last 7 days, sourced from OpenRouter public rankings.

Sources:OpenRouter Usage

#	Model	OpenRouter Usage	AISOTA Score
1SOTA	deepseek-v4-flash	917.01M	99.8
22	gpt-5-6-luna	361.6M	99.5
33	glm-5-3-flash	334.96M	99.3
4	gemini-2-5-flash-lite	292.09M	99.1
5	gpt-oss-20b	131.11M	98.9
6	gemini-3-1-flash-lite	126.55M	98.6
7	text-embedding-3-small	124.66M	98.4
8	gemini-2-5-flash	123.6M	98.2
9	gpt-oss-120b	106.15M	97.9
10	gpt-4o-mini	106.01M	97.7

AI Tools Leaderboard

Ranked lists of the best AI tools across the most popular categories, with editorial notes on each tool.

AI Business Tools

318 tools

1SOTAFigma99.7
2Rose AI99.4
3Asana99.1
4Comment Generator98.7
5Sprout Social98.4

AI Code Tools

76 tools

AI Productivity Tools

233 tools

1SOTAFrame99.6
2Sonix99.1
3Everlaw98.7
4Torq98.3
5Poised97.8

AI Text Generators

128 tools

1SOTAStable Diffusion Prompt Generator99.2
2Cohesive98.4
3StudyX97.7
4ProWritingAid96.9
5Rankify AI96.1

AI Image Tools

183 tools

AI Audio Generators

91 tools

1SOTASplice98.9
2iZotope RX97.8
3S10.AI96.7
4Wondercraft95.6
5Gladia94.4

AI Video Tools

76 tools

AI Art Generators

65 tools

Automation Tools

81 tools

1SOTADot98.8
2Zoho Sheet97.5
3Simple Phones96.3
4Humans95.1
5Nabla93.8

Misc AI Tools

42 tools

AI Compute & API Pricing

Live AI compute prices: model API inference costs, cloud GPU benchmark pricing, and best value models per quality point. Updated daily from public sources.

API Inference Pricing

2761 entries

1SOTAdeepinfra/meta-llama/Llama-3.2-3B-Instruct$0.02/M
2lambda_ai/llama3.2-3b-instruct$0.02/M
3nscale/Qwen/Qwen2.5-Coder-3B-Instruct$0.02/M
4nscale/Qwen/Qwen2.5-Coder-7B-Instruct$0.02/M
5nebius/Qwen/Qwen2.5-Coder-7B$0.02/M

Chinese AI Models Pricing

869 entries

1SOTAnscale/Qwen/Qwen2.5-Coder-3B-Instruct$0.02/M
2nscale/Qwen/Qwen2.5-Coder-7B-Instruct$0.02/M
3nebius/Qwen/Qwen2.5-Coder-7B$0.02/M
4nscale/deepseek-ai/DeepSeek-R1-Distill-Llama-8B$0.03/M
5novita/qwen/qwen3-4b-fp8$0.03/M

Best Value Models (per quality point)

11 entries

1SOTAdeepseek-v4-pro$0.100
2claude-sonnet-5$0.110
3deepseek-v4-flash$0.117
4claude-opus-5$0.168
5claude-opus-4-8$0.241

Cloud GPU Benchmark Pricing

18 entries

1SOTAStandard_NV4as_v4$0.054/h
2Standard_NC36ds_xl_RTXPRO6000BSE_v6$0.333/h
3Standard_NV24s_v3$0.584/h
4Standard_NV16as_v4$1.901/h
5Standard_NC320lds_xl_RTXPRO6000BSE_v6$2.643/h