GPT-4: Benchmarks, Pricing & Context Window
GPT-4 is a language model from OpenAI, released in June 2023, with multimodal input.
GPT-4 is a large multimodal model capable of processing both image and text inputs and generating human-like text outputs. It demonstrates human-level performance on various professional and academic benchmarks.
GPT-4 benchmarks
Capability tiers
Standing within each category, adjusted for leaderboard depth.
Real tasks performance
High-confidence performance for GPT-4 across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.
Performance by conversation depth
How GPT-4 holds up as conversations get longer.
Quality Tracker
GPT-4 Performance Across Datasets
Scores sourced from the model's scorecard, paper, or official blog posts
GPT-4 pricing
Providers
GPT-4 starts at $30.00 per million input tokens and $60.00 per million output tokens via Azure. See all 2 providers below with their per-token pricing, latency, throughput, and modality support.
| Provider | Input $/M | Cached input $/M | Output $/M | Context in / out | TTFT p95 s | Output p5 c/s | Modalities in / out |
|---|---|---|---|---|---|---|---|
| $30.00 | — | $60.00 | 32.8K/32.8K | 0.30 | — | / | |
| $30.00 | — | $60.00 | 32.8K/32.8K | 0.50 | — | / |
Cached input is the discounted price for prompt tokens served from a provider cache. TTFT is time to first token. Output is characters per second; p5 is the sustained floor exceeded by 95% of observed requests.
GPT-4 context window
Input and output token limits for GPT-4, plus how it ranks on long-context understanding.
Try now
Make it with
GPT-4.
GPT-4 latency
GPT-4 time to first token, sustained output throughput, and failed-request rate from live model usage over the trailing 7 days.
GPT-4 examples
Recent arena outputs from GPT-4, picked from the highest-ranked matchups.
GPT-4 license
GPT-4 is a proprietary model available under its provider's product and API terms, has a knowledge cutoff of December 2022.
- License
- Proprietary
- Hosted access
- Knowledge cutoff
- December 2022
Proprietary license - usage restrictions apply
GPT-4 resources
Official sources for GPT-4: provider documentation, official playground, paper or system card, official launch post, source repository.
GPT-4 vs other models
The most-compared alternatives to GPT-4 are Claude 3 Opus, GPT-4 Turbo, Grok-2 mini. Open any pair side-by-side for benchmarks, pricing, context, and latency.
Models like GPT-4
Models ranked just above and below GPT-4 by LLM Stats score.
FAQ
Common questions about GPT-4.