This model is deprecated. We still benchmark its performance, but less often: every 18 hours instead of every 6.
OpenAI has launched a newer release, GPT-4o (Nov '24). We suggest considering it instead.
GPT-4 Turbo API Provider Benchmarking & Analysis
This analysis is intended to support you in choosing the best API provider of GPT-4 Turbo for your use-case.
Lowest Time per Task
Time per Intelligence Index task (minutes)
Total 1 provider
Lowest Cost per Task
Cost per Intelligence Index task
Total 1 provider
GPT-4 Turbo is available through Azure.
Pricing
Cost per Intelligence Index Task: GPT-4 Turbo
Pricing: Cache Hit, Input, and Output: GPT-4 Turbo
Output Speed vs. Cost per Task: GPT-4 Turbo
Speed
Average of the single-request and multi-turn agentic workloads measured
Output Speed: GPT-4 Turbo
Latency
Average of the single-request and multi-turn agentic workloads measured
Time to First Answer Token: GPT-4 Turbo
Reliability
Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured
Endpoint Error Rate: GPT-4 Turbo
Caching
Cache Hit Rate: GPT-4 Turbo
Cost per Task vs. Cache Hit Rate: GPT-4 Turbo
Key Comparison Metrics & API Features
Frequently Asked Questions
Common questions about GPT-4 Turbo providers
GPT-4 Turbo is available through 1 API provider: Azure. Each provider offers different performance characteristics and pricing.
GPT-4 Turbo is currently available through 1 API provider that we benchmark and track.
The provider with the lowest time to first answer token for GPT-4 Turbo is Azure (1.18s). Lower latency means faster initial response time.
The most affordable provider for GPT-4 Turbo by blended price is Azure ($12.00 per 1M tokens). Blended price uses a 7:2:1 cache hit/input/output token ratio.
The provider with the lowest input token pricing for GPT-4 Turbo is Azure ($10.00 per 1M input tokens).
The provider with the lowest output token pricing for GPT-4 Turbo is Azure ($30.00 per 1M output tokens).
The only provider of GPT-4 Turbo supports JSON mode for structured output.
The only provider of GPT-4 Turbo supports function calling (tool use).
When choosing a provider for GPT-4 Turbo, consider: time per task (how quickly the provider completes an Intelligence Index task), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and API features like JSON mode or function calling.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
For information about GPT-4 Turbo's intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview