36 models

DeepSeek Models: Intelligence, Performance & Price

Name: Artificial Analysis Intelligence Index
Creator: Artificial Analysis
License: https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf

DeepSeek

Intelligence

DeepSeek V4.1 Flash (max)

DeepSeek V4 Pro 0813 (max)

DeepSeek V4 Flash Vision (max)

DeepSeek V4.1 Flash (Non-reasoning)

Artificial Analysis Intelligence Index

Speed

DeepSeek V4.1 Flash (Non-reasoning)

239 t/s

DeepSeek V4 Flash Vision (max)

223 t/s

DeepSeek V4.1 Flash (max)

221 t/s

DeepSeek V4 Pro 0813 (max)

104 t/s

Output speed

Cost

DeepSeek V4.1 Flash (Non-reasoning)

$0.15

DeepSeek V4.1 Flash (max)

$0.27

DeepSeek V4 Flash Vision (max)

$0.31

DeepSeek V4 Pro 0813 (max)

$0.67

Per Intelligence Index Task

Artificial Analysis has benchmarked 36 models from DeepSeek. Below is a comparison of the key metrics across these models.

For intelligence, the top model from DeepSeek is DeepSeek V4.1 Flash (Reasoning, Max Effort) at 39.
For output speed, the fastest model is DeepSeek V4.1 Flash (Non-Reasoning) at 239 t/s.
For latency, DeepSeek V4.1 Flash (Non-Reasoning) at 1.12s offers the lowest time to first answer token.
For pricing, DeepSeek V4.1 Flash (Non-Reasoning) at $0.15 offers the lowest cost per task. Prices vary up to 4.6x across models.

Intelligence

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1

Artificial Analysis Intelligence Index v4.3.2 includes: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Intelligence Index vs. Cost per Intelligence Index Task

Artificial Analysis Intelligence Index · Weighted average cost (USD) per Artificial Analysis Intelligence Index task

Most attractive quadrant

Pareto line

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.

Cost

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better

Speed & Latency

Output Speed

Output tokens per second · Higher is better

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).

Figures represent performance of the model's first-party API or the median across providers where a first-party API is not available.

Capability Scores

Capability Indexes

Measures the performance of models on specific capabilities and industries

Finance & Accounting Index

Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

Strategy & Ops Index

Incorporates 6 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

Legal Index

Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AA-LCR v1.1, GDP.pdf, AutomationBench-AA · Higher is better

Healthcare & Medical Index

Incorporates 6 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, MLCR-AA, Humanity's Last Exam, AutomationBench-AA · Higher is better

Engineering Index

Incorporates 6 evaluations: AA-Omniscience, Humanity's Last Exam, CritPt, GDPval-AA v2.1, AA-Briefcase v1.1, Terminal-Bench 4.0 · Higher is better

Economics Index

Incorporates 5 evaluations: AA-Omniscience, Humanity's Last Exam, GDPval-AA v2.1, AA-Briefcase v1.1, AA-LCR v1.1 · Higher is better

All DeepSeek Releases

Further details

							Weights		Provider Benchmarks
DeepSeek V4.1 Flash (Reasoning, Max Effort)	DeepSeek	39	552B 16B active at inference time	1M	$0.2	221	🤗	+15	View
DeepSeek V4 Pro 0813 (Reasoning, Max Effort)	DeepSeek	36	1.6T 49B active at inference time	1M	$0.7	104	🤗	+8	View
DeepSeek V4 Flash Vision (Reasoning, Max Effort)	DeepSeek	35	284B 13B active at inference time	1M	$0.2	223	Not available	+2	View
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)	DeepSeek	34	284B 13B active at inference time	1M	$0.2	227	🤗	+14	View
DeepSeek V4 Pro 0424 (Reasoning, Max Effort)	DeepSeek	30	1.6T 49B active at inference time	1M	$0.2	111	🤗	+7	View
DeepSeek V4 Pro 0424 (Reasoning, High Effort)	DeepSeek	30	1.6T 49B active at inference time	1M	$0.2	106	🤗	+5	View
DeepSeek V4 Flash 0420 (Reasoning, High Effort)	DeepSeek	26	284B 13B active at inference time	1M	$0.1	-	🤗	+4	View
DeepSeek V4.1 Flash (Non-Reasoning)	DeepSeek	25	552B 16B active at inference time	1M	$0.2	239	🤗		View
DeepSeek V4 Flash 0420 (Reasoning, Max Effort)	DeepSeek	24	284B 13B active at inference time	1M	$0.1	-	🤗	+2	View
DeepSeek V3.2 (Reasoning)	DeepSeek	21	685B 37B active at inference time	128k	$0.1	-	🤗	+5	View
DeepSeek V4 Pro 0424 (Non-reasoning)	DeepSeek	21	1.6T 49B active at inference time	1M	$0.2	99	🤗	+3	View
DeepSeek V4 Flash 0420 (Non-reasoning)	DeepSeek	19	284B 13B active at inference time	1M	$0.1	-	🤗		View
DeepSeek V3.2 Exp (Reasoning)	DeepSeek	17	685B 37B active at inference time	128k	$0.1	-	🤗		View
DeepSeek V3.2 (Non-reasoning)	DeepSeek	16	685B 37B active at inference time	128k	$0.3	-	🤗	+8	View
DeepSeek V3.1 Terminus (Reasoning)	DeepSeek	15	685B 37B active at inference time	128k	$1.7	-	🤗		View
DeepSeek V3.2 Speciale	DeepSeek	14	685B 37B active at inference time	128k	-	-	🤗	-	View
DeepSeek V3.1 Terminus (Non-reasoning)	DeepSeek	14	685B 37B active at inference time	128k	$0.2	-	🤗		View
DeepSeek V3.2 Exp (Non-reasoning)	DeepSeek	14	685B 37B active at inference time	128k	$0.1	-	🤗		View
DeepSeek V3.1 (Non-reasoning)	DeepSeek	14	685B 37B active at inference time	128k	$0.7	-	🤗	+5	View
DeepSeek V3.1 (Reasoning)	DeepSeek	13	685B 37B active at inference time	128k	$0.7	-	🤗		View
DeepSeek R1 0528 (May '25)	DeepSeek	13	685B 37B active at inference time	128k	$1.5	-	🤗	+2	View
DeepSeek R1 (Jan '25)	DeepSeek	11	685B 37B active at inference time	128k	$2.2	-	🤗		View
DeepSeek V3 0324	DeepSeek	10	671B 37B active at inference time	128k	$0.8	-	🤗		View
DeepSeek V3 (Dec '24)	DeepSeek	8	671B 37B active at inference time	128k	$0.4	-	🤗		View
DeepSeek R1 Distill Qwen 32B	DeepSeek	8	32B	128k	-	-	🤗	-	View
DeepSeek R1 0528 Qwen3 8B	DeepSeek	8	8.2B	33k	-	-	🤗	-	View
DeepSeek R1 Distill Llama 70B	DeepSeek	8	70B	128k	$0.7	-	🤗		View
DeepSeek R1 Distill Qwen 14B	DeepSeek	8	14B	128k	-	-	🤗	-	View
DeepSeek-V2.5 (Dec '24)	DeepSeek	7	236B 21B active at inference time	128k	-	-	🤗	-	View
DeepSeek-V2.5	DeepSeek	7	236B 21B active at inference time	128k	-	-	🤗	-	View
DeepSeek R1 Distill Llama 8B	DeepSeek	6	8B	128k	-	-	🤗	-	View
DeepSeek-Coder-V2	DeepSeek	6	236B 21B active at inference time	128k	-	-	🤗	-	View
DeepSeek R1 Distill Qwen 1.5B	DeepSeek	6	1.5B	128k	-	-	🤗	-	View
DeepSeek-V2-Chat	DeepSeek	6	236B 21B active at inference time	128k	-	-	🤗	-	View
DeepSeek Coder V2 Lite Instruct	DeepSeek	5	16B 2.4B active at inference time	128k	-	-	🤗	-	View
DeepSeek LLM 67B Chat (V1)	DeepSeek	5	7B	4k	-	-	🤗	-	View

DeepSeek Models: Intelligence, Performance & Price

Intelligence

Speed

Cost

Intelligence

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index

Intelligence Index vs. Cost per Intelligence Index Task

Cost per Intelligence Index Task

Artificial Analysis Intelligence Index

Cost

Cost per Intelligence Index Task

Cost per Intelligence Index Task

Speed & Latency

Output Speed

Output Speed

Model Performance Representation

Capability Scores

Capability Indexes

All DeepSeek Releases

Further details