AWS Bedrock Pricing
AWS Bedrock charges from $0.040 per 1 million output tokens, depending on the model. Voxtral Mini 3B 2507 is the cheapest at $0.040/1M output and $0.040/1M input; its current flagship Claude Opus 4.1 (2025-08-05) costs $75.00/1M output. The largest context window is 10.0M tokens (Llama4 Scout 17B Instruct). Prices are USD, current as of October 12, 2026.
Bedrock is a multi-vendor surface — Anthropic, Meta, Mistral, Cohere and Amazon Nova models all billed through AWS on-demand token pricing. Provisioned throughput is priced per model-unit-hour instead and is excluded from the on-demand rates below.
Last updated · synced weekly from the upstream model catalog
Cheapest input
$0.035/1M
Nova Micro
Cheapest output
$0.040/1M
Voxtral Mini 3B 2507
Longest context
10.0M
Llama4 Scout 17B Instruct
Models priced
131
Avg $10.96/1M output
Pricing by model
All prices in USD per 1 million tokens. Showing 80 of 131 priced models, cheapest first. The batch column is the asynchronous batch-API rate, 50% below the standard output rate on the 7 models that publish one.
| Model | Input / 1M | Output / 1M | Batch out / 1M | Context |
|---|---|---|---|---|
| Voxtral Mini 3B 2507 | $0.040 | $0.040 | — | 33K |
| Gemma 3 4B It | $0.040 | $0.080 | — | 128K |
| Ministral 3 3B Instruct | $0.100 | $0.100 | — | 128K |
| Llama3 2 1B Instruct | $0.130 | $0.130 | — | 128K |
| Nova Micro | $0.035 | $0.140 | — | 128K |
| Ministral 3 8B Instruct | $0.150 | $0.150 | — | 128K |
| Llama3 2 11B Instruct | $0.160 | $0.160 | — | 128K |
| Llama3 2 3B Instruct | $0.190 | $0.190 | — | 128K |
| GPT Oss Safeguard 20B | $0.070 | $0.200 | — | 128K |
| Ministral 3 14B Instruct | $0.200 | $0.200 | — | 128K |
| Mistral 7B Instruct | $0.150 | $0.200 | — | 32K |
| Titan Text Lite | $0.150 | $0.200 | — | 42K |
| Llama3 1 8B Instruct | $0.220 | $0.220 | — | 128K |
| Nemotron Nano 9B | $0.060 | $0.230 | — | 128K |
| Nemotron Nano 3 30B | $0.060 | $0.240 | — | 256K |
| Nova Lite | $0.060 | $0.240 | — | 300K |
| Us Gov.nvidia.nemotron Nano 9B | $0.072 | $0.276 | — | 128K |
| Us Gov.nvidia.nemotron Nano 3 30B | $0.072 | $0.288 | — | 262K |
| Gemma 3 12B It | $0.090 | $0.290 | — | 128K |
| GPT Oss 20B 1 | $0.070 | $0.300 | — | 128K |
| Voxtral Small 24B 2507 | $0.100 | $0.300 | — | 33K |
| Us Gov.openai.gpt Oss 20B 1 | $0.084 | $0.360 | — | 128K |
| Gemma 3 27B It | $0.230 | $0.380 | — | 128K |
| Glm 4.7 Flash | $0.070 | $0.400 | — | 203K |
| Jamba 1.5 Mini | $0.200 | $0.400 | — | 256K |
| Claude Haiku 5.5 | $0.100 | $0.500 | — | 1.0M |
| GPT 6 Luna | $0.110 | $0.550 | — | 1.1M |
| Command Light Text | $0.300 | $0.600 | — | 4K |
| GPT Oss 120B 1 | $0.150 | $0.600 | — | 128K |
| GPT Oss Safeguard 120B | $0.150 | $0.600 | — | 128K |
| Llama3 8B Instruct | $0.300 | $0.600 | — | 8K |
| Nemotron Nano 12B | $0.200 | $0.600 | — | 128K |
| Palmyra Vision 7B | $0.150 | $0.600 | — | 4K |
| Qwen3 32B | $0.150 | $0.600 | — | 33K |
| Qwen3 Coder 30B A3b | $0.150 | $0.600 | — | 262K |
| Titan Text Express | $0.200 | $0.600 | — | 42K |
| Us Gov.anthropic.claude Haiku 5.5 | $0.120 | $0.600 | — | 1.0M |
| Nemotron Super 3 120B | $0.150 | $0.650 | — | 256K |
| Llama4 Scout 17B Instruct | $0.170 | $0.660 | $0.330−50% | 10.0M |
| Jamba Instruct | $0.500 | $0.700 | — | 70K |
| Mixtral 8X7B Instruct | $0.450 | $0.700 | — | 32K |
| Llama3 1 70B Instruct | $0.720 | $0.720 | — | 128K |
| Llama3 2 90B Instruct | $0.720 | $0.720 | — | 128K |
| Llama3 3 70B Instruct | $0.720 | $0.720 | — | 128K |
| Us Gov.nvidia.nemotron Nano 12B | $0.240 | $0.720 | — | 128K |
| Us Gov.openai.gpt Oss 120B 1 | $0.180 | $0.720 | — | 128K |
| Us Gov.nvidia.nemotron Super 3 120B | $0.180 | $0.780 | — | 256K |
| Qwen3 235B A22b 2507 | $0.220 | $0.880 | $0.440−50% | 262K |
| Llama4 Maverick 17B Instruct | $0.240 | $0.970 | $0.485−50% | 1.0M |
| Llama2 13B Chat | $0.750 | $1.00 | — | 4K |
| Minimax M2 | $0.300 | $1.20 | — | 1.0M |
| Minimax M2.1 | $0.300 | $1.20 | — | 196K |
| Minimax M2.5 | $0.300 | $1.20 | — | 196K |
| Qwen3 Coder Next | $0.500 | $1.20 | — | 256K |
| Qwen3 Next 80B A3b | $0.150 | $1.20 | — | 256K |
| GPT 5.6 Luna | $0.220 | $1.32 | — | 1.1M |
| Magistral Small 2509 | $0.500 | $1.50 | — | 128K |
| Mistral Large 3 675B Instruct | $0.500 | $1.50 | — | 256K |
| Titan Text Premier | $0.500 | $1.50 | — | 42K |
| V3 | $0.580 | $1.68 | — | 128K |
| Qwen3 Coder 480B A35b | $0.450 | $1.80 | — | 131K |
| V3.2 | $0.620 | $1.85 | — | 164K |
| Command Text | $1.00 | $2.00 | — | 4K |
| Devstral 2 123B | $0.400 | $2.00 | — | 256K |
| Glm 4.7 | $0.600 | $2.20 | — | 203K |
| Claude Instant | $0.800 | $2.40 | — | 100K |
| Llama3 1 405B Instruct | $2.40 | $2.40 | — | 128K |
| Kimi K2 Thinking | $0.600 | $2.50 | — | 256K |
| Nova 2 Lite | $0.300 | $2.50 | — | 1.0M |
| Llama2 70B Chat | $1.95 | $2.56 | — | 4K |
| Qwen3 VL 235B A22b | $0.530 | $2.66 | — | 256K |
| Kimi K2.5 | $0.600 | $3.00 | — | 256K |
| Mistral Small 2402 | $1.00 | $3.00 | $1.50−50% | 32K |
| Glm 5 | $1.00 | $3.20 | — | 200K |
| Nova Pro | $0.800 | $3.20 | — | 300K |
| Llama3 70B Instruct | $2.65 | $3.50 | — | 8K |
| Claude 3.5 Haiku (2024-10-22) | $0.800 | $4.00 | — | 200K |
| Claude Haiku 4.5 (2025-10-01) | $1.00 | $5.00 | $2.50−50% | 200K |
| Glm 5.3 | $1.68 | $5.28 | $2.64−50% | 1.0M |
| Claude Mythos Preview | $27.50 | $137.50 | — | 1.0M |
A dash in the batch column means our catalog carries no published batch rate for that model — not that the model has no batch tier. Batch coverage in the upstream pricing data is uneven.
Frequently asked questions
How much does AWS Bedrock cost per 1M tokens?
AWS Bedrock pricing starts at $0.040 per 1 million output tokens (Voxtral Mini 3B 2507) across 131 models, with its current flagship Claude Opus 4.1 (2025-08-05) at $75.00. Older premium models in the catalog list higher. Rates current as of October 12, 2026.
What is the cheapest AWS Bedrock model?
Voxtral Mini 3B 2507 is the cheapest AWS Bedrock model on output tokens at $0.040 per 1M, while Nova Micro is cheapest on input at $0.035 per 1M. Which wins for you depends on your input-to-output ratio.
What is the largest AWS Bedrock context window?
Llama4 Scout 17B Instruct has the largest context window in the AWS Bedrock catalog at 10.0M input tokens, with up to 8K output tokens per response.
Does AWS Bedrock offer batch pricing?
Yes — 7 AWS Bedrock models in this catalog publish a discounted rate for asynchronous batch jobs, at 50% off output tokens. Claude Haiku 4.5 (2025-10-01) bills $2.50 per 1M output in batch against $5.00 synchronously. Batch trades real-time responses for the lower rate, so it suits backfills, evaluations and bulk enrichment rather than user-facing calls.
How do I calculate my actual AWS Bedrock bill?
Multiply your input tokens by the input rate and your output tokens by the output rate, both per million, then add retries and any conversation history resent on each turn. Calcaas does this against your real usage pattern and shows the margin left at your price point.
Other providers
What will this actually cost you?
Per-token rates only tell you half the story. Calcaas multiplies them by your real usage — prompt size, response length, retries and conversation history — then shows the margin left at your price point.
Open the free cost calculator