The most common strategic question in AI projects is whether to build a custom model or use a foundation model. The framing has evolved: it used to be “build vs. buy a pre-trained model”; it is now “fine-tune a custom model vs. use a large foundation model with prompting.” The right answer depends on your data situation, volume, accuracy requirements, and team capability.

Foundation Models via Bedrock

Foundation models available through Amazon Bedrock are trained on massive datasets and perform well on a wide range of tasks out of the box. Bedrock has grown into a multi-provider catalog: Anthropic Claude, Amazon Nova (the successor to the earlier Amazon Titan family), Meta Llama, Mistral, Cohere, and, since June 2026, OpenAI models (GPT-5.5, GPT-5.4, and Codex are generally available on Bedrock). You access them via API. The default on-demand pricing is per token (you pay separately for input and output tokens), with batch inference, provisioned throughput, and prompt caching as lower-cost options for the right workloads.

Advantages:

  • No training data required - works immediately with prompting
  • No model training cost or time - operational from day one
  • Regular capability improvements from provider without your effort
  • Handles novel tasks and edge cases via instruction following

Disadvantages:

  • Per-token cost at scale - costs grow linearly with volume. As an illustration, processing on the order of 1 million tokens per day at a mid-tier model rate (for example a Claude Sonnet tier) lands roughly in the low thousands of EUR per month. Check the current Bedrock pricing page for exact per-token rates, which vary by model, provider, and region.
  • Latency - LLM inference is slower than purpose-built classifiers (100-500ms vs. 10-50ms)
  • Output variability - generative models produce different outputs for the same input; determinism requires additional engineering
  • Data privacy - your data is sent to a third-party API (Bedrock uses AWS infrastructure, but the data leaves your compute environment)

Best fit: Tasks requiring language understanding, generation, or reasoning. Varied or unpredictable inputs. Low to medium volume (under 100,000 complex calls/day). Situations where flexibility matters more than optimized cost.

Custom ML on SageMaker AI

The classic model building service is now called Amazon SageMaker AI. The original Amazon SageMaker was renamed SageMaker AI at re:Invent 2024, and the name Amazon SageMaker now refers to the broader unified data, analytics, and AI platform (with SageMaker Unified Studio at its center). SageMaker AI remains the place you build, train, and deploy models, available standalone or inside the unified platform.

Training a purpose-built model on your own labeled data produces a model specialized for your specific task. SageMaker AI is not only for classic ML, it also supports fine-tuning and customizing foundation models (for example custom Amazon Nova models), so the line between the two columns below is more of a spectrum than a wall.

Advantages:

  • Lower inference cost at scale - a deployed SageMaker endpoint for a small classifier costs 100-300 EUR/month regardless of volume
  • Lower latency - small models respond in 5-20ms
  • Full data control - training and inference run entirely in your AWS environment
  • Deterministic outputs - classification models return consistent results
  • Optimized accuracy for your specific use case when sufficient data exists

Disadvantages:

  • Requires labeled training data - typically 1,000+ examples per class minimum, 10,000+ for strong performance
  • Training time and cost - a moderately complex model takes hours to days to train and costs 50-500 EUR in compute
  • Maintenance burden - models drift as input distribution changes; retraining is ongoing work
  • Limited generalization - a custom classifier does exactly what it was trained for, nothing else
  • Team capability requirement - building and maintaining SageMaker pipelines requires ML engineering skills

Best fit: High-volume classification, regression, or structured prediction tasks. Well-defined problem with a stable schema. Sufficient labeled training data. Team with ML engineering capability.

The Hybrid Architecture

Most mature AI systems use both. The common pattern:

  1. Foundation model during prototyping and for complex/rare cases
  2. Foundation model generates training data labels for high-volume tasks
  3. Custom model trained on those labels handles the volume
  4. Foundation model serves as fallback when custom model confidence is low

This gives cost efficiency at scale while maintaining flexibility for edge cases. The custom model handles 80-90% of traffic at low cost; the foundation model handles the remaining 10-20% where it is needed.

Decision Framework

FactorPoints toward foundation modelPoints toward custom model
Training data availableNoYes (1,000+/class)
Daily volumeLow (<10,000 calls)High (>100,000 calls)
Task definitionFuzzy, reasoning-intensiveWell-defined, structured
Latency requirement>200ms acceptable<50ms required
Team ML capabilityLimitedStrong
TimelineWeeksMonths

For a head to head on the two services themselves, see /comparisons/sagemaker-vs-bedrock/. The build versus buy decision more broadly is covered in /comparisons/build-vs-buy-ai/, and fine-tuning versus prompting in /comparisons/fine-tuning-vs-prompt-engineering/.

Sources