Artificial Analysis Intelligence Index v4.1 includes: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR . See Intelligence Index methodology：/methodology/intelligence-benchmarking for further details, including a breakdown of each evaluation and how we run them.
Indicates whether the model weights are available. Models are labelled as 'Commercial Use Restricted' if the weights are available but commercial use is limited (typically requires obtaining a paid license).
Agentic real-world work tasks, (Elo-500)/2000
Legal agentic work, criterion pass rate
Kubernetes incident root-cause analysis
While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.
AA-Briefcase Elo is a combined metric that aggregates analytical quality Elo, presentation Elo, and rubric pass rate, with rubric performance converted into Elo via synthetic head-to-head matches. Elo and 95% confidence interval bounds are clamped at 0.
AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct.
Weighted average cost per Intelligence Index task. Each evaluation’s cost is calculated from input, cache hit, cache write, reasoning, and answer token prices, divided by task count, and weighted by its Intelligence Index weight.
The number of tokens required per Intelligence Index task. This is calculated by multiplying the output tokens per eval by the relative weights of each benchmark in the Intelligence Index, then dividing by task count (excluding repeats).
The cost to run the evaluations in the Artificial Analysis Intelligence Index, calculated using the model's input, cache hit, cache write, reasoning, and answer token prices and the number of tokens used across evaluations (excluding repeats).
Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.
Price per token included in the request/message sent to the API, represented as USD per million Tokens.
The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:
See Prompt Caching：/models/caching for the full breakdown.
Price per token generated by the model (received from the API), represented as USD per million Tokens.
Figures represent performance of the model's first-party API (e.g. OpenAI for o1) or the median across providers where a first-party API is not available (e.g. Meta's Llama models).
Larger context windows are relevant to RAG (Retrieval Augmented Generation) LLM workflows which typically involve reasoning and information retrieval of large amounts of data.
Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).
Measured by Output Speed (tokens per second)
Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).
The weighted average time (seconds) per Artificial Analysis Intelligence Index task. This is calculated by dividing output tokens per task by output speed, weighted by the relative weights of each benchmark in the Intelligence Index.
Measured by Time (seconds) to First Token
Time to first answer token received, in seconds, after API request sent. For reasoning models, this includes the 'thinking' time of the model before providing an answer. For models which do not support streaming, this represents time to receive the completion.
Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed
Seconds to receive a 500 token response. Key components:
The total number of trainable weights and biases in the model, expressed in billions. These parameters are learned during training and determine the model's ability to process and generate responses.
The number of parameters actually executed during each inference forward pass, expressed in billions. For Mixture of Experts (MoE) models, a routing mechanism selects a subset of experts per token, resulting in fewer active than total parameters. Dense models use all parameters, so active equals total.
Claude Opus 5 (Adaptive Reasoning, Max Effort) currently leads the Artificial Analysis Intelligence Index with a score of 61, out of 170 models evaluated.
The top AI models by Intelligence Index are: 1. Claude Opus 5 (Adaptive Reasoning, Max Effort) (61), 2. Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) (60), 3. Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) (60), 4. GPT-5.6 Sol (max) (59), and 5. Claude Opus 5 (Adaptive Reasoning, High Effort) (59).
Mercury 2 is the fastest at 901.6 tokens per second, followed by HyperNova 60B 2605 (414.2 t/s) and Granite 4.0 H Small (407.1 t/s).
Nova Micro is the most affordable at $0.03 per 1M tokens (blended), followed by Sarvam 30B (high) ($0.03) and Gemma 4 E4B (Non-reasoning) ($0.03).
Gemini 2.5 Flash-Lite (Non-reasoning) has the lowest time to first token at 0.35s, followed by Command A+ (0.42s) and Gemini 2.5 Flash (Non-reasoning) (0.49s).
GLM-5.2 (max) is the highest-ranked open weights model with an Intelligence Index score of 51. There are 94 open weights models out of 170 total evaluated.
The top open weights AI models by Intelligence Index are: 1. GLM-5.2 (max) (51), 2. MiniMax-M3 (44), and 3. DeepSeek V4 Pro (Reasoning, Max Effort) (44).
Claude Opus 5 (Adaptive Reasoning, Max Effort) leads among 126 reasoning models with an Intelligence Index score of 61. Reasoning models use extended thinking to work through complex problems before providing answers.
Models are compared across multiple dimensions including intelligence (quality), pricing, output speed (tokens per second), latency (time to first token), end-to-end response time, and context window size. Performance metrics are measured directly using standardized prompts across 586 models.
Click on any model name or row in the charts to view its dedicated page with detailed metrics and direct comparisons against similar models. You can also use the model selector to customize which models appear in each chart. View the leaderboard：/leaderboards/models
Scan to open this article