Salesforce Models API Models — Pricing & Benchmarks
10 models available · Salesforce
Last refreshed 2026-10-17. Next refresh: weekly.
Salesforce Models API hosts 10 AI models in this catalog. Per-token pricing is not listed for these Salesforce Models API rows yet; compare context windows, benchmarks, and hosting options instead. LLM Reference compares these models across all 134 providers.
| Model | Input (per 1M) | Output (per 1M) | Context | |
|---|---|---|---|---|
| Claude 3 Haiku | — | — | 200K | |
| Claude Sonnet 5 | — | — | 1M | |
| GPT-3.5 Turbo | — | — | 16K | |
| GPT-3.5 Turbo (Instruct) | — | — | 4K | |
| GPT-3.5 Turbo 16k | — | — | 16K | |
| GPT-4 | — | — | 8K | |
| GPT-4 Turbo | — | — | 128K | |
| GPT-4o (08-06) | — | — | 128K | |
| GPT-5 | — | — | 400K | |
| GPT-5.5 | — | — | 1.05M |
Where else to run this
GPT-3.5 Turbo 16k on Salesforce Models API
Provider setup and pricing
GPT-4 on Salesforce Models API
Provider setup and pricing
Claude 3 Haiku on Salesforce Models API
Provider setup and pricing
GPT-3.5 Turbo 16k on Azure OpenAI
Alternative host
GPT-4 on OpenAI API
Alternative host
Claude 3 Haiku on AWS Bedrock
Alternative host
About Salesforce Models API
The Models API sends prompts from Apex or REST to Salesforce-managed models or to bring-your-own models. Requests pass through the Einstein Trust Layer, and with data masking on, context is limited to 65,536 tokens. Each call consumes Einstein Requests: total input and output tokens divided by 2,000 and rounded up, then multiplied by a usage-type multiplier.
Full provider profile →