Amazon Bedrock model customisation vs Tinker

Amazon Bedrock model customisation

Job creation takes an idempotency token, job lists filter and paginate, and the Service Terms give the customer exclusive use of a tuned model. Jobs run in two US Regions only, weights can't be exported, and the newest customisation API change found dates from 28 May 2026.

Tinker

Full control of the training loop with the GPUs abstracted away, plus recipes for SFT, DPO, RL and distillation. LoRA only; no full-parameter training.

Amazon Bedrock model customisation

1. Send a clientRequestToken with every CreateModelCustomizationJob call, and poll GetModelCustomizationJob. Jobs are asynchronous and can take hours
2. Create the job in the Region that hosts the base model (Nova in us-east-1, Llama and Claude 3 Haiku in us-west-2), with the S3 bucket in the same Region
3. Pass an IAM service role that trusts bedrock.amazonaws.com and can read the training data and write the output location. The caller needs iam:PassRole
4. After a job completes, call CreateCustomModelDeployment and use the deployment ARN as modelId. Models outside the on-demand list need Provisioned Throughput
5. For gpt-oss-20b and Qwen3 32B, use /v1/fine_tuning/jobs on bedrock-mantle.us-west-2.api.aws with a Bedrock API key and a Lambda grader ARN

Tinker

1. Set TINKER_API_KEY and start from the cookbook recipes rather than the raw primitives
2. Read the 'Avoid Client-Side Timeouts and Retries' guide before wrapping sampling calls in your own retries; the SDK already retries sampling with stable request IDs
3. Save intermediate checkpoints with a TTL between 1 hour and 10 years; storage bills at $0.10 a GB-month until they expire
4. Read models.json for current prices before a run; sampling tokens cost more than training tokens on the open models
5. Check the model deprecations page before pinning a base model; 18 were retired on 2026-06-12