Best Fine Tuning Companies

ComparisonChecked Nov 8, 2026

Google Cloud Vertex AI vs Amazon Bedrock for fine-tuning

Both tune models inside a hyperscaler account, with the tuned model served in that cloud. They differ in the model families on offer, the methods per model, and published pricing.

What differs

  • Models: Vertex AI tunes Gemini models and open models such as Gemma 3, Llama, and Qwen 3; Bedrock fine-tunes Amazon Nova, Meta Llama 3.x, and Claude 3 Haiku, with model-specific regions.
  • Methods: Vertex AI lists SFT, preference tuning (Gemini 2.5 Flash and Flash-Lite), and RL fine-tuning (Gemini 3.5 Flash and 3.1 Flash-Lite); Bedrock lists SFT, reinforcement fine-tuning with Lambda reward functions, and distillation.
  • Pricing: Vertex AI publishes per-token tuning prices (for example Gemini 2.5 Flash $0.005 per 1,000 training tokens); Bedrock publishes RFT at $80.00 per training hour for gpt-oss-20b and Qwen3 32B plus $1.95 monthly storage.
  • Weights: neither profile lists a path to export tuned closed-model weights.
  • Published security: Google lists Generative AI on Agent Platform in SOC 2, ISO/IEC 27001, and HIPAA scope; AWS states Bedrock is in scope for ISO and SOC programs, HIPAA eligible, and FedRAMP High in GovCloud (US-West).

Who each tends to fit

The choice usually follows the cloud your data and security reviews already sit in, and which model family you need to tune. Check method support for your exact model first.

Last checked: . No winner declared.

Side by side

FactGoogle Cloud Vertex AI tuning (Agent Platform)Amazon Bedrock (model customization)
StatusActive. Business line of Google Cloud; parent Alphabet Inc. is publicly listed.Active. Amazon Bedrock is a business line of Amazon Web Services; parent Amazon.com, Inc. is publicly listed.
OwnershipPublicly listedPublicly listed
Buying modelManaged fine-tuning API, Hyperscaler cloud platformManaged fine-tuning API, Hyperscaler cloud platform
CategoriesManaged fine-tuning API; Cloud platform fine-tuning; Reinforcement fine-tuningManaged fine-tuning API; Cloud platform fine-tuning; Reinforcement fine-tuning
PricingSupervised fine-tuning per 1,000 training tokens: Gemini 2.5 Pro $0.025; Gemini 3.5 Flash $0.01 (SFT and RL fine-tuning); Gemini 2.5 Flash $0.005; Gemini 3.5 Flash Lite and 3.1 Flash Lite $0.003; Gemini 2.5 Flash Lite $0.0015. Open models per 1M training tokens, for example Llama 3.1 8B $0.67, Gemma 3 27B IT $6.83, Qwen 3 32B $6.57. Training tokens = dataset tokens x epochs.Reinforcement fine-tuning: $80.00 per training hour for gpt-oss-20b and Qwen3 32B (US East N. Virginia and US West Oregon), plus $1.95 storage per trained model per month. Supervised fine-tuning is billed per token processed (training tokens x epochs) plus monthly storage; rates vary by model.
CertificationsSOC 2 (Generative AI on Agent Platform in scope); ISO/IEC 27001 (in scope); HIPAA (Generative AI on Agent Platform listed; BAA required)In scope for AWS ISO and SOC compliance programs (as stated by AWS); HIPAA eligible; GDPR (customers can use Bedrock in compliance with GDPR, per AWS); FedRAMP High (AWS GovCloud US-West)
LimitationsTuned Gemini models are served in Google Cloud; Gemini weights are not exportable; From Gemini 3, tuned model endpoint prediction price is 1.5x the base model in non-global regions (per pricing page); Method support differs by model; check the tuning table before planningFine-tuning model list is limited and region-specific (for example, Nova models in us-east-1, Llama models in us-west-2); No weights-export path found in the Bedrock customization docs; custom models run inside Bedrock; Supervised fine-tuning rates for current models were not readable on the public pricing page when checked; confirm in the AWS pricing calculator
How to buycloud.google.comaws.amazon.com
Sources[1] [2] [3] [4] [5] Checked 2026-11-08[1] [2] [3] [4] Checked 2026-11-08

Facts come from each vendor’s profile sources. Run a paid pilot before choosing.