Skip to main content
Migration is intentionally boring. In every language, the change is confined to imports and client construction — plus swapping Bedrock model ARNs for AI Reserve model IDs (for example gpt-5.4-mini, gpt-5.5, claude-opus-4-8). Message shapes, tool configs, streaming loops, and error handling carry over verbatim.
Java note. converseStream in the AI Reserve SDK is synchronous (a blocking push handler or a pull iterator) rather than the AWS async CompletableFuture publisher — for most codebases this removes boilerplate rather than adding it.

Bedrock-routed models

Some teams prefer a model’s serving behavior via AWS Bedrock over the provider’s direct API. AI Reserve offers both paths behind the same endpoint: append -bedrock to a model ID and the gateway routes that request through AWS Bedrock instead of the provider directly. Everything else — SDK, key, shapes, streaming, errors — is identical, so you can A/B the two paths by changing only the model string.
Models that Bedrock does not serve (OpenAI GPT, Gemini, Grok, Perplexity Sonar, Moonshot) have no -bedrock variant — the direct IDs remain the only path for those. Every variant routes through AWS Bedrock. Since October 7, 2026 the Claude variants come in two geographies, priced exactly as Bedrock prices them: Append -bedrock-us instead of -bedrock when your data must stay in the United States (for example claude-opus-4-8-bedrock-us). The twelve Claude variants above all have a -bedrock-us twin; the Llama, Mistral and DeepSeek variants are single-leg. If your code sends a raw AWS id, the us. form resolves to the -bedrock-us leg and the global. / bare forms resolve to the -bedrock leg (see AWS SDK).

Latency

Compared to calling a provider (or Bedrock) directly from your own code, the AI Reserve gateway adds a small, fixed toll — authentication, rate limiting, quota checks, and request translation: Rule of thumb: ~60–200 ms added to time-to-first-token, and effectively zero between tokens — streaming responses are piped through without re-buffering, so inter-token latency is determined entirely by the model provider. Against typical generation times of seconds to tens of seconds, the overhead is low single-digit percent. The toll does not grow with context size or generation length. Note that provider-side differences (for example, Bedrock vs. direct-API time-to-first-token for the same Claude model, which varies by region and load) are independent of the gateway and typically larger than the gateway’s own overhead. The -bedrock variants make measuring both paths for your workload a one-string A/B test. Already on boto3 or the AWS JS SDK and want to keep them? See AWS SDK (boto3) — the gateway serves the Bedrock Converse API at the same paths, so unmodified clients work by overriding the endpoint.