¿Amazon Bedrock ofrece modelos de pesos abiertos?

Sí, 43 de 95 modelos en Amazon Bedrock son modelos de pesos abiertos: Kimi K2.5, GLM-4.7, Kimi K2 Thinking, DeepSeek V3.2, MiniMax-M2, GLM-4.7, DeepSeek V3.2, gpt-oss-120b (high), GLM-4.7-Flash, Qwen3 Coder Next, DeepSeek V3.1, DeepSeek V3.1, DeepSeek R1 (Jan), Qwen3 235B 2507, Qwen3 Coder 480B, Mistral Large 3, GLM-4.7-Flash, gpt-oss-120b (low), gpt-oss-20b (high), gpt-oss-20b (low), Llama 4 Maverick, Qwen3 Coder 30B A3B, Magistral Small 1.2, Ministral 3 14B, Llama 4 Scout, Llama 3.3 70B, Ministral 3 8B, Qwen3 32B, Llama 3.1 8B, Gemma 3 27B, NVIDIA Nemotron Nano 9B V2, Mistral Large 2 (Jul), Llama 3.1 70B Standard, Llama 3.1 70B Latency Optimized, Ministral 3 3B, Gemma 3 12B, Jamba 1.5 Large, NVIDIA Nemotron Nano 12B v2 VL, Command-R+ (Apr), Jamba 1.5 Mini, Mixtral 8x7B, Command-R (Mar) y Gemma 3 4B.

Artificial Analysis

Amazon Bedrock: inteligencia, rendimiento y precio de sus modelos

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

Amazon Bedrock

Análisis de los modelos de Amazon Bedrock en métricas clave como calidad, precio, velocidad de salida, latencia, ventana de contexto y más. Este análisis está pensado para ayudarte a elegir el mejor modelo ofrecido por Amazon Bedrock para tu caso de uso.

Más inteligente

Claude Opus 5 (max)

Claude Opus 5 (xhigh)

GPT-5.6 Sol (max)

Claude Opus 5 (high)

Claude Opus 5 (medium)

Índice de inteligencia

95 modelos en total

Más rápido

Ministral 3 3B

402 t/s

Nova Micro

276 t/s

Ministral 3 8B

244 t/s

GPT-5.6 Luna (max)

234 t/s

Llama 4 Maverick

221 t/s

Velocidad de salida

95 modelos en total

Menor precio

Nova Micro

$0.03

Gemma 3 4B

$0.04

Nova Lite

$0.05

NVIDIA Nemotron Nano 9B V2

$0.08

gpt-oss-20b (high)

$0.09

Precio combinado (por 1M de tokens)

95 modelos en total

Indica un modelo de razonamiento

Amazon ofrece 95 modelos, cada uno con distintas características de inteligencia, rendimiento y precio. A continuación se comparan las métricas clave entre modelos.

En inteligencia, los mejores modelos en Amazon son Claude Opus 5 (max) (61), Claude Opus 5 (xhigh) (60) y GPT-5.6 Sol (max) (59).
En velocidad de salida, los modelos más rápidos son Ministral 3 3B (402 t/s), Nova Micro (276 t/s) y Ministral 3 8B (244 t/s). La velocidad varía significativamente entre modelos, con una diferencia del 82% entre el más rápido y el más lento.
En latencia, GPT-5.4 (Non-reasoning) (0.57s), GPT-5.5 (Non-reasoning) (0.70s) y Grok 4.3 (Non-reasoning) (0.73s) ofrecen el menor tiempo hasta el primer token de respuesta.
En precios, Nova Micro ($0.03), Gemma 3 4B ($0.04) y Nova Lite ($0.05) ofrecen los menores precios combinados por 1M de tokens. Los precios varían hasta 3.4x entre modelos.
En tamaño de ventana de contexto, Claude Opus 5 (max) (1M), Claude Opus 5 (xhigh) (1M) y Claude Opus 5 (high) (1M) admiten las ventanas de contexto más grandes en Amazon.

Aspectos destacados

Inteligencia

Artificial Analysis Intelligence Index · Higher is better

Velocidad

Output tokens per second · Higher is better

Precio

USD per 1M tokens (blended) · Lower is better

Evaluaciones de inteligencia

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR

Reasoning models are indicated by a lightbulb icon

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 for further details, including a breakdown of each evaluation and how we run them.

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better

GDPval-AA v2

Agentic real-world work tasks, (Elo-500)/2000

𝜏³-Banking

Agentic tool use

Terminal-Bench v2.1

Agentic coding & terminal use

SciCode

Coding

Humanity's Last Exam

Reasoning & knowledge

GPQA Diamond

Scientific reasoning

CritPt

Physics reasoning

AA-Omniscience Accuracy

Knowledge

AA-Omniscience Non-Hallucination Rate

1 - hallucination rate

AA-LCR

Long context reasoning

AA-Briefcase

Agentic knowledge work, Elo

AutomationBench-AA

Agentic SaaS workflows

Harvey LAB-AA

Legal agentic work, criterion pass rate

EnterpriseOps-Gym-AANew

Agentic business operations

IFBench

Instruction following

APEX-Agents-AA

Long-horizon agentic tasks

ITBench-AA

Kubernetes incident root-cause analysis

MMMU-Pro

Visual reasoning

Reasoning models are indicated by a lightbulb icon

While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.

Intelligence Index vs. Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)

Most attractive quadrant

Reasoning models are indicated by a lightbulb icon

While higher intelligence models are typically more expensive, they do not all follow the same price-quality curve.

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:

Anthropic: charges a separate cache write fee, with different rates for 5-minute and 1-hour TTLs (1-hour TTL is more expensive).
Google (Vertex/Gemini): charges a per-hour cache storage fee in addition to cache hit pricing. Some providers also use tiered pricing for prompts above 200K tokens.
OpenAI, DeepSeek, others: typically charge only cache hit pricing with no write or storage fee.

See Prompt Caching for the full breakdown.

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).

Ventana de contexto

Context Window

Context window: tokens limit · Higher is better

Reasoning models are indicated by a lightbulb icon

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).

Precios

Intelligence Index vs. Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)

Most attractive quadrant

Reasoning models are indicated by a lightbulb icon

While higher intelligence models are typically more expensive, they do not all follow the same price-quality curve.

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

The blended cache price shown here uses cache hit price only. Other caching costs differ by provider:

Anthropic: charges a separate cache write fee, with different rates for 5-minute and 1-hour TTLs (1-hour TTL is more expensive).
Google (Vertex/Gemini): charges a per-hour cache storage fee in addition to cache hit pricing. Some providers also use tiered pricing for prompts above 200K tokens.
OpenAI, DeepSeek, others: typically charge only cache hit pricing with no write or storage fee.

See Prompt Caching for the full breakdown.

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).

Resumen de rendimiento

Output Speed vs. Price

Output speed: output tokens per second · USD per 1M tokens (blended)

Most attractive quadrant

Reasoning models are indicated by a lightbulb icon

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).

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Velocidad

Medida por la velocidad de salida (tokens por segundo)

Output Speed

Output tokens per second · Higher is better

Reasoning models are indicated by a lightbulb icon

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 (e.g. OpenAI for o1) or the median across providers where a first-party API is not available (e.g. Meta's Llama models).

Latencia

Medida por el tiempo (segundos) hasta el primer token

Latency: Time To First Answer Token

Seconds to first answer token received · Accounts for reasoning model 'thinking' time

Reasoning models are indicated by a lightbulb icon

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.

Tiempo de respuesta de extremo a extremo

Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed

End-to-End Response Time vs. Price

End-to-end response time: end-to-end seconds to output 500 tokens · USD per 1M tokens (blended)

Most attractive quadrant

Reasoning models are indicated by a lightbulb icon

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Seconds to receive a 500 token response. Key components:

Input time: Time to receive the first response token
Thinking time (only for reasoning models): Time reasoning models spend outputting tokens to reason prior to providing an answer. Amount of tokens based on the average reasoning tokens across a diverse set of 60 prompts (methodology details).
Answer time: Time to generate 500 output tokens, based on output speed

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.


									Análisis adicional
Claude Opus 5 (max)	1M	Propietario	61	$2.03	54	45.57	54.91	--	Modelo Proveedor
Claude Opus 5 (xhigh)	1M	Propietario	60	$1.56	55	30.93	39.95	--	Modelo Proveedor
GPT-5.6 Sol (max)	262k	Propietario	59	$1.39	132	113.40	117.20	--	Modelo Proveedor
Claude Opus 5 (high)	1M	Propietario	59	$1.06	53	19.43	28.93	--	Modelo Proveedor
Claude Opus 5 (medium)	1M	Propietario	56	$0.62	51	7.20	16.95	--	Modelo Proveedor
Claude Opus 4.8 (max)	1M	Propietario	56	$1.78	58	21.18	29.87	--	Modelo Proveedor
GPT-5.6 Terra (max)	262k	Propietario	55	$0.83	166	151.21	154.21	--	Modelo Proveedor
GPT-5.5 (xhigh)	272k	Propietario	55	$1.82	129	63.18	67.06	--	Modelo Proveedor
Claude Opus 4.7 (max)	1M	Propietario	54	$3.26	62	1.61	9.68	--	Modelo Proveedor
Claude Sonnet 5 (max)	1M	Propietario	53	$1.73	64	216.36	224.17	--	Modelo Proveedor
GPT-5.5 (high)	272k	Propietario	53	$1.22	108	14.79	19.44	--	Modelo Proveedor
GPT-5.4 (xhigh)	272k	Propietario	51	$1.57	171	114.68	117.61	--	Modelo Proveedor
GPT-5.6 Luna (max)	262k	Propietario	51	$0.30	234	125.83	127.97	--	Modelo Proveedor
Claude Opus 5 (low)	1M	Propietario	51	$0.36	52	3.38	13.01	--	Modelo Proveedor
GPT-5.5 (medium)	272k	Propietario	50	$0.74	109	7.70	12.28	--	Modelo Proveedor
Claude Sonnet 4.6 (max)	200k	Propietario	47	$1.21	53	101.62	111.06	--	Modelo Proveedor
Claude Opus 4.6 (max)	1M	Propietario	44*	--	48	6.77	17.18	--	Modelo Proveedor
GPT-5.5 (low)	272k	Propietario	43	$0.38	108	1.43	6.07	--	Modelo Proveedor
Claude Opus 4.7 (Non-reasoning, high)	1M	Propietario	43*	--	58	1.53	10.16	--	Modelo Proveedor
Claude Sonnet 5 (Non-reasoning)	1M	Propietario	42	$0.54	67	1.96	9.47	--	Modelo Proveedor
GPT-5.6 Sol (Non-reasoning)	262k	Propietario	41	$0.25	93	4.02	9.39	--	Modelo Proveedor
Claude Opus 4.5	200k	Propietario	41*	--	46	13.28	24.03	--	Modelo Proveedor
GPT-5.4 (low)	272k	Propietario	39*	--	130	1.11	4.96	--	Modelo Proveedor
Claude Opus 4.6 (high)	1M	Propietario	38*	--	46	1.82	12.68	--	Modelo Proveedor
Grok 4.3 (high)	524k	Propietario	38	$0.37	186	18.67	21.36	--	Modelo Proveedor
Claude 4.5 Sonnet	1M	Propietario	36	$0.42	52	8.91	18.54	--	Modelo Proveedor
Grok 4.3 (medium)	524k	Propietario	36*	--	187	8.34	11.01	--	Modelo Proveedor
Claude Sonnet 4.6 (Non-reasoning)	200k	Propietario	36*	--	48	1.36	11.88	--	Modelo Proveedor
Grok 4.3 (low)	524k	Propietario	35*	--	172	3.33	6.23	--	Modelo Proveedor
GPT-5.5 (Non-reasoning)	272k	Propietario	35	$0.31	106	0.70	5.41	--	Modelo Proveedor
Kimi K2.5	256k	Abierto	35	$0.23	89	1.34	40.33	33.37	Modelo Proveedor
Claude Opus 4.5	200k	Propietario	35*	--	46	1.85	12.82	--	Modelo Proveedor
GPT-5.6 Terra (Non-reasoning)	262k	Propietario	34	$0.18	138	0.73	4.35	--	Modelo Proveedor
Claude 4.1 Opus	200k	Propietario	34*	--	--	--	--	--	Modelo Proveedor
GLM-4.7	200k	Abierto	34	$0.32	67	1.65	39.23	30.07	Modelo Proveedor
Kimi K2 Thinking	256k	Abierto	33*	--	122	1.44	21.96	16.41	Modelo Proveedor
DeepSeek V3.2	128k	Abierto	32	$0.22	57	1.46	45.14	34.94	Modelo Proveedor
Claude 4.5 Haiku	200k	Propietario	30	$0.22	111	9.88	14.39	--	Modelo Proveedor
Claude 4.5 Sonnet	1M	Propietario	29*	--	49	1.74	11.95	--	Modelo Proveedor
Claude 4 Sonnet	1M	Propietario	29	$0.44	--	--	--	--	Modelo Proveedor
MiniMax-M2	205k	Abierto	28*	--	122	1.06	21.50	16.35	Modelo Proveedor
Claude 4.1 Opus	200k	Propietario	28*	--	--	--	--	--	Modelo Proveedor
GPT-5.4 (Non-reasoning)	272k	Propietario	28*	--	130	0.57	4.42	--	Modelo Proveedor
GPT-5.6 Luna (Non-reasoning)	262k	Propietario	27	$0.09	214	2.99	5.33	--	Modelo Proveedor
GLM-4.7	200k	Abierto	27*	--	68	1.55	8.86	--	Modelo Proveedor
Claude 4 Sonnet	1M	Propietario	25*	--	--	--	--	--	Modelo Proveedor
Grok 4.3 (Non-reasoning)	524k	Propietario	25	$0.99	179	0.73	3.53	--	Modelo Proveedor
DeepSeek V3.2	128k	Abierto	25*	--	97	1.45	6.60	--	Modelo Proveedor
gpt-oss-120b (high)	131k	Abierto	24	$0.06	92	1.07	28.29	21.77	Modelo Proveedor
Claude 4.5 Haiku	200k	Propietario	24*	--	102	1.00	5.90	--	Modelo Proveedor
GLM-4.7-Flash	200k	Abierto	23*	--	183	1.22	14.89	10.94	Modelo Proveedor
Nova 2.0 Pro Preview (medium)	256k	Propietario	22	$0.17	114	12.75	34.69	17.55	Modelo Proveedor
Qwen3 Coder Next	128k	Abierto	21	$0.46	134	2.15	5.88	--	Modelo Proveedor
DeepSeek V3.1	128k	Abierto	21*	--	174	1.43	4.31	--	Modelo Proveedor
Nova 2.0 Omni (medium)	1M	Propietario	21*	--	--	--	--	--	Modelo Proveedor
Nova 2.0 Omni (medium)	1M	Propietario	21*	--	--	--	--	--	Modelo Proveedor
DeepSeek V3.1	128k	Abierto	21*	--	168	1.43	16.27	11.87	Modelo Proveedor
Nova 2.0 Pro Preview (low)	256k	Propietario	20	$0.21	115	10.28	31.93	17.32	Modelo Proveedor
Nova 2.0 Lite (medium)	1M	Propietario	19*	--	150	19.24	35.95	13.37	Modelo Proveedor
DeepSeek R1 (Jan)	128k	Abierto	19	$0.22	167	1.13	18.14	14.02	Modelo Proveedor
Nova 2.0 Lite (high)	1M	Propietario	18	$0.25	--	--	--	--	Modelo Proveedor
Nova 2.0 Lite (high)	1M	Propietario	18	$0.25	151	22.60	39.17	13.26	Modelo Proveedor
Qwen3 235B 2507	256k	Abierto	18*	--	59	1.36	9.79	--	Modelo Proveedor
Qwen3 Coder 480B	262k	Abierto	18*	--	91	1.27	6.76	--	Modelo Proveedor
Nova 2.0 Lite (low)	1M	Propietario	18*	--	142	9.50	27.10	14.08	Modelo Proveedor
Nova 2.0 Omni (low)	1M	Propietario	17*	--	--	--	--	--	Modelo Proveedor
Mistral Large 3	256k	Abierto	16	$0.06	172	1.18	4.08	--	Modelo Proveedor
GLM-4.7-Flash	200k	Abierto	16*	--	198	1.13	3.66	--	Modelo Proveedor
gpt-oss-120b (low)	131k	Abierto	15	$0.02	91	1.07	28.55	21.99	Modelo Proveedor
gpt-oss-20b (high)	131k	Abierto	15	$0.03	90	52.33	80.11	22.23	Modelo Proveedor
Nova 2.0 Pro Preview	256k	Propietario	14	$0.25	92	1.06	6.51	--	Modelo Proveedor
gpt-oss-20b (low)	131k	Abierto	14*	--	73	13.54	47.63	27.27	Modelo Proveedor
Llama 4 Maverick	128k	Abierto	14	$0.03	221	0.91	3.18	--	Modelo Proveedor
Qwen3 Coder 30B A3B	262k	Abierto	14*	--	187	1.15	3.82	--	Modelo Proveedor
Nova Premier	1M	Propietario	13*	--	32	2.88	18.61	--	Modelo Proveedor
Nova 2.0 Lite	1M	Propietario	12*	--	136	1.15	4.84	--	Modelo Proveedor
Magistral Small 1.2	128k	Abierto	11	$0.25	74	1.59	35.54	27.16	Modelo Proveedor
Ministral 3 14B	256k	Abierto	11	$0.15	181	1.13	3.88	--	Modelo Proveedor
Nova 2.0 Omni	1M	Propietario	11*	--	--	--	--	--	Modelo Proveedor
Llama 4 Scout	128k	Abierto	10	$0.01	162	0.90	3.98	--	Modelo Proveedor
Claude 3.5 Sonnet (Oct)	200k	Propietario	10*	--	--	--	--	--	Modelo Proveedor
Llama 3.3 70B	128k	Abierto	9	$0.10	119	1.34	5.52	--	Modelo Proveedor
Ministral 3 8B	256k	Abierto	9	$0.18	244	1.04	3.09	--	Modelo Proveedor
Qwen3 32B	32.8k	Abierto	9*	--	154	1.19	4.43	--	Modelo Proveedor
Claude 3.5 Sonnet (June)	200k	Propietario	8*	--	--	--	--	--	Modelo Proveedor
Nova Pro	300k	Propietario	8*	--	105	1.14	5.91	--	Modelo Proveedor
Llama 3.1 8B	128k	Abierto	8	--	172	0.75	3.66	--	Modelo Proveedor
Gemma 3 27B	128k	Abierto	7	$0.19	63	1.49	9.39	--	Modelo Proveedor
NVIDIA Nemotron Nano 9B V2	131k	Abierto	7*	--	169	1.13	4.10	--	Modelo Proveedor
Mistral Large 2 (Jul)	128k	Abierto	7*	--	30	1.80	18.24	--	Modelo Proveedor
Nova Lite	300k	Propietario	7*	--	153	0.95	4.22	--	Modelo Proveedor
Llama 3.1 70B Standard	128k	Abierto	7*	--	96	1.36	6.56	--	Modelo Proveedor
Llama 3.1 70B Latency Optimized	128k	Abierto	7*	--	119	1.24	5.46	--	Modelo Proveedor
Ministral 3 3B	256k	Abierto	6	$0.13	402	0.97	2.21	--	Modelo Proveedor
Gemma 3 12B	128k	Abierto	6	$0.13	97	1.39	6.56	--	Modelo Proveedor
Jamba 1.5 Large	256k	Abierto	5*	--	46	2.76	13.63	--	Modelo Proveedor
Claude 3 Sonnet	200k	Propietario	5*	--	45	3.08	14.18	--	Modelo Proveedor
Nova Micro	130k	Propietario	5*	--	276	0.93	2.74	--	Modelo Proveedor
NVIDIA Nemotron Nano 12B v2 VL	128k	Abierto	5*	--	198	1.17	3.69	--	Modelo Proveedor
Mistral Large (Feb)	32.8k	Propietario	4*	--	42	1.57	13.52	--	Modelo Proveedor
Claude 3 Haiku	200k	Propietario	4*	--	76	0.98	7.52	--	Modelo Proveedor
Llama 3 70B	8.19k	Abierto	3*	--	--	--	--	--	Modelo Proveedor
Command-R+ (Apr)	128k	Abierto	3*	--	37	2.47	16.10	--	Modelo Proveedor
Jamba 1.5 Mini	256k	Abierto	3*	--	53	4.53	14.04	--	Modelo Proveedor
Mixtral 8x7B	32.8k	Abierto	2*	--	24	1.15	22.11	--	Modelo Proveedor
Mistral 7B	8.19k	Abierto	2*	--	--	--	--	--	Modelo Proveedor
Command-R (Mar)	128k	Abierto	2*	--	69	1.31	8.60	--	Modelo Proveedor
Llama 3 8B	8.19k	Abierto	1*	--	--	--	--	--	Modelo Proveedor
Gemma 3 4B	128k	Abierto	1*	--	176	1.07	3.92	--	Modelo Proveedor

Definiciones clave

Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).

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).

Time to first token received, in seconds, after API request sent. For reasoning models which share reasoning tokens, this will be the first reasoning token. For models which do not support streaming, this represents time to receive the completion.

Average cost per task in the index. Costs are split by input, cache hit, cache write, reasoning, and answer token pricing where canonical token counts are available.

Price per token included in the request/message sent to the API, represented as USD per million Tokens.

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 to write prompt tokens into the cache so that later requests can hit them, represented as USD per million tokens. Some providers charge a premium over the standard input price to create a cache entry (e.g. Anthropic), while others cache automatically with no separate write fee.

Price per token generated by the model (received from the API), represented as USD per million Tokens.

Metrics are 'live' and are based on the past 72 hours of measurements, measurements are taken 8 times a day for single requests and 2 times per day for parallel requests.

Preguntas frecuentes

Preguntas comunes sobre Amazon Bedrock

Amazon Bedrock ofrece 95 modelos que monitoreamos: Claude Opus 5 (max), Claude Opus 5 (xhigh), GPT-5.6 Sol (max), Claude Opus 5 (high), Claude Opus 5 (medium), Claude Opus 4.8 (max), GPT-5.6 Terra (max), GPT-5.5 (xhigh), Claude Opus 4.7 (max), Claude Sonnet 5 (max), GPT-5.5 (high), GPT-5.4 (xhigh), GPT-5.6 Luna (max), Claude Opus 5 (low), GPT-5.5 (medium), Claude Sonnet 4.6 (max), Claude Opus 4.6 (max), GPT-5.5 (low), Claude Opus 4.7 (Non-reasoning, high), Claude Sonnet 5 (Non-reasoning), GPT-5.6 Sol (Non-reasoning), Claude Opus 4.5, GPT-5.4 (low), Claude Opus 4.6 (high), Grok 4.3 (high), Claude 4.5 Sonnet, Grok 4.3 (medium), Claude Sonnet 4.6 (Non-reasoning), Grok 4.3 (low), GPT-5.5 (Non-reasoning), Kimi K2.5, Claude Opus 4.5, GPT-5.6 Terra (Non-reasoning), GLM-4.7, Kimi K2 Thinking, DeepSeek V3.2, Claude 4.5 Haiku, Claude 4.5 Sonnet, MiniMax-M2, GPT-5.4 (Non-reasoning), GPT-5.6 Luna (Non-reasoning), GLM-4.7, Grok 4.3 (Non-reasoning), DeepSeek V3.2, gpt-oss-120b (high), Claude 4.5 Haiku, GLM-4.7-Flash, Nova 2.0 Pro Preview (medium), Qwen3 Coder Next, DeepSeek V3.1, DeepSeek V3.1, Nova 2.0 Pro Preview (low), Nova 2.0 Lite (medium), DeepSeek R1 (Jan), Nova 2.0 Lite (high), Qwen3 235B 2507, Qwen3 Coder 480B, Nova 2.0 Lite (low), Mistral Large 3, GLM-4.7-Flash, gpt-oss-120b (low), gpt-oss-20b (high), Nova 2.0 Pro Preview, gpt-oss-20b (low), Llama 4 Maverick, Qwen3 Coder 30B A3B, Nova Premier, Nova 2.0 Lite, Magistral Small 1.2, Ministral 3 14B, Llama 4 Scout, Llama 3.3 70B, Ministral 3 8B, Qwen3 32B, Nova Pro, Llama 3.1 8B, Gemma 3 27B, NVIDIA Nemotron Nano 9B V2, Mistral Large 2 (Jul), Nova Lite, Llama 3.1 70B Standard, Llama 3.1 70B Latency Optimized, Ministral 3 3B, Gemma 3 12B, Jamba 1.5 Large, Claude 3 Sonnet, Nova Micro, NVIDIA Nemotron Nano 12B v2 VL, Mistral Large (Feb), Claude 3 Haiku, Command-R+ (Apr), Jamba 1.5 Mini, Mixtral 8x7B, Command-R (Mar) y Gemma 3 4B.

El modelo más inteligente disponible en Amazon Bedrock es Claude Opus 5 (max), con una puntuación de 61 en el Índice de Inteligencia.

El modelo más rápido en Amazon Bedrock por velocidad de salida es Ministral 3 3B, con 402.1 tokens por segundo.

El modelo con el menor tiempo hasta el primer token de respuesta en Amazon Bedrock es GPT-5.4 (Non-reasoning), con 0.57s. Una menor latencia significa una respuesta inicial más rápida.

El modelo más económico en Amazon Bedrock por precio combinado es Nova Micro, a $0.03 por 1M de tokens (proporción 7:2:1 de aciertos de caché/entrada/salida).

Los precios en Amazon Bedrock varían hasta 177x entre modelos, desde $0.03 por 1M de tokens para Nova Micro hasta $4.80 por 1M de tokens para Mistral Large (Feb).

22 de 95 modelos en Amazon Bedrock admiten el modo JSON para salida estructurada.

93 de 95 modelos en Amazon Bedrock admiten llamadas a funciones (uso de herramientas).

Sí, Amazon Bedrock ofrece 43 modelos de razonamiento: Claude Opus 5 (max), Claude Opus 5 (xhigh), GPT-5.6 Sol (max), Claude Opus 5 (high), Claude Opus 5 (medium), Claude Opus 4.8 (max), GPT-5.6 Terra (max), GPT-5.5 (xhigh), Claude Opus 4.7 (max), Claude Sonnet 5 (max), GPT-5.5 (high), GPT-5.4 (xhigh), GPT-5.6 Luna (max), Claude Opus 5 (low), GPT-5.5 (medium), Claude Sonnet 4.6 (max), Claude Opus 4.6 (max), GPT-5.5 (low), Claude Opus 4.5, GPT-5.4 (low), Grok 4.3 (high), Claude 4.5 Sonnet, Grok 4.3 (medium), Grok 4.3 (low), Kimi K2.5, GLM-4.7, Kimi K2 Thinking, DeepSeek V3.2, Claude 4.5 Haiku, MiniMax-M2, gpt-oss-120b (high), GLM-4.7-Flash, Nova 2.0 Pro Preview (medium), DeepSeek V3.1, Nova 2.0 Pro Preview (low), Nova 2.0 Lite (medium), DeepSeek R1 (Jan), Nova 2.0 Lite (high), Nova 2.0 Lite (low), gpt-oss-120b (low), gpt-oss-20b (high), gpt-oss-20b (low) y Magistral Small 1.2. Los modelos de razonamiento usan pensamiento extendido para resolver problemas complejos antes de responder.

Sí, el rendimiento de un proveedor puede variar con el tiempo debido a cambios de infraestructura, balanceo de carga y actualizaciones. Medimos continuamente a todos los proveedores y mostramos las tendencias históricas de rendimiento en los gráficos "a lo largo del tiempo".

Al elegir un modelo en Amazon Bedrock, considera: inteligencia (para tareas sensibles a la calidad), velocidad de salida (para tareas intensivas en rendimiento), latencia (para aplicaciones interactivas que requieren primeras respuestas rápidas), precios (para cargas de trabajo sensibles al costo) y funcionalidades como el tamaño de la ventana de contexto, el modo JSON o las llamadas a funciones.

Amazon Bedrock: inteligencia, rendimiento y precio de sus modelos

Más inteligente

Más rápido

Menor precio

Inteligencia

Velocidad

Precio